FR Doc 2010-8159

[Federal Register: May 7, 2010 (Volume 75, Number 88)]
[Rules and Regulations]
[Page 25323-25728]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr07my10-8]

[[Page 25323]]

-----------------------------------------------------------------------

Part II

Environmental Protection Agency

-----------------------------------------------------------------------

Department of Transportation

-----------------------------------------------------------------------

National Highway Traffic Safety Administration

40 CFR Parts 85, 86, and 600; 49 CFR Parts 531, 533, 536, et al.

Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate
Average Fuel Economy Standards; Final Rule

[[Page 25324]]

-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 85, 86, and 600

DEPARTMENT OF TRANSPORTATION

National Highway Traffic Safety Administration

49 CFR Parts 531, 533, 536, 537 and 538

[EPA-HQ-OAR-2009-0472; FRL-9134-6; NHTSA-2009-0059]
RIN 2060-AP58; RIN 2127-AK50

Light-Duty Vehicle Greenhouse Gas Emission Standards and
Corporate Average Fuel Economy Standards; Final Rule

AGENCY: Environmental Protection Agency (EPA) and National Highway
Traffic Safety Administration (NHTSA).

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: EPA and NHTSA are issuing this joint Final Rule to establish a
National Program consisting of new standards for light-duty vehicles
that will reduce greenhouse gas emissions and improve fuel economy.
This joint Final Rule is consistent with the National Fuel Efficiency
Policy announced by President Obama on May 19, 2009, responding to the
country's critical need to address global climate change and to reduce
oil consumption. EPA is finalizing greenhouse gas emissions standards
under the Clean Air Act, and NHTSA is finalizing Corporate Average Fuel
Economy standards under the Energy Policy and Conservation Act, as
amended. These standards apply to passenger cars, light-duty trucks,
and medium-duty passenger vehicles, covering model years 2012 through
2016, and represent a harmonized and consistent National Program. Under
the National Program, automobile manufacturers will be able to build a
single light-duty national fleet that satisfies all requirements under
both programs while ensuring that consumers still have a full range of
vehicle choices. NHTSA's final rule also constitutes the agency's
Record of Decision for purposes of its National Environmental Policy
Act (NEPA) analysis.

DATES: This final rule is effective on July 6, 2010, sixty days after
date of publication in the Federal Register. The incorporation by
reference of certain publications listed in this regulation is approved
by the Director of the Federal Register as of July 6, 2010.

ADDRESSES: EPA and NHTSA have established dockets for this action under
Docket ID No. EPA-HQ-OAR-2009-0472 and NHTSA-2009-0059, respectively.
All documents in the docket are listed on the http://
www.regulations.gov Web site. Although listed in the index, some
information is not publicly available, e.g., CBI or other information
whose disclosure is restricted by statute. Certain other material, such
as copyrighted material, is not placed on the Internet and will be
publicly available only in hard copy form. Publicly available docket
materials are available either electronically through http://
www.regulations.gov or in hard copy at the following locations: EPA:
EPA Docket Center, EPA/DC, EPA West, Room 3334, 1301 Constitution Ave.,
NW., Washington, DC. The Public Reading Room is open from 8:30 a.m. to
4:30 p.m., Monday through Friday, excluding legal holidays. The
telephone number for the Public Reading Room is (202) 566-1744. NHTSA:
Docket Management Facility, M-30, U.S. Department of Transportation,
West Building, Ground Floor, Rm. W12-140, 1200 New Jersey Avenue, SE.,
Washington, DC 20590. The Docket Management Facility is open between 9
a.m. and 5 p.m. Eastern Time, Monday through Friday, except Federal
holidays.

FOR FURTHER INFORMATION CONTACT:
EPA: Tad Wysor, Office of Transportation and Air Quality,
Assessment and Standards Division, Environmental Protection Agency,
2000 Traverwood Drive, Ann Arbor MI 48105; telephone number: 734-214-
4332; fax number: 734-214-4816; e-mail address: wysor.tad@epa.gov, or
Assessment and Standards Division Hotline; telephone number (734) 214-
4636; e-mail address asdinfo@epa.gov. NHTSA: Rebecca Yoon, Office of
Chief Counsel, National Highway Traffic Safety Administration, 1200 New
Jersey Avenue, SE., Washington, DC 20590. Telephone: (202) 366-2992.

SUPPLEMENTARY INFORMATION:

Does this action apply to me?

This action affects companies that manufacture or sell new light-
duty vehicles, light-duty trucks, and medium-duty passenger vehicles,
as defined under EPA's CAA regulations,\1\ and passenger automobiles
(passenger cars) and non-passenger automobiles (light trucks) as
defined under NHTSA's CAFE regulations.\2\ Regulated categories and
entities include:
---------------------------------------------------------------------------

\1\ ``Light-duty vehicle,'' ``light-duty truck,'' and ``medium-
duty passenger vehicle'' are defined in 40 CFR 86.1803-01.
Generally, the term ``light-duty vehicle'' means a passenger car,
the term ``light-duty truck'' means a pick-up truck, sport-utility
vehicle, or minivan of up to 8,500 lbs gross vehicle weight rating,
and ``medium-duty passenger vehicle'' means a sport-utility vehicle
or passenger van from 8,500 to 10,000 lbs gross vehicle weight
rating. Medium-duty passenger vehicles do not include pick-up
trucks.
\2\ ``Passenger car'' and ``light truck'' are defined in 49 CFR
part 523.

------------------------------------------------------------------------
Examples of potentially
Category NAICS codes \A\ regulated entities
------------------------------------------------------------------------
Industry................. 336111, 336112..... Motor vehicle
manufacturers.
Industry................. 811112, 811198, Commercial Importers of
541514. Vehicles and Vehicle
Components.
------------------------------------------------------------------------
\A\North American Industry Classification System (NAICS).

This list is not intended to be exhaustive, but rather provides a
guide regarding entities likely to be regulated by this action. To
determine whether particular activities may be regulated by this
action, you should carefully examine the regulations. You may direct
questions regarding the applicability of this action to the person
listed in FOR FURTHER INFORMATION CONTACT.

Table of Contents

I. Overview of Joint EPA/NHTSA National Program
A. Introduction
1. Building Blocks of the National Program
2. Public Participation
B. Summary of the Joint Final Rule and Differences From the
Proposal
1. Joint Analytical Approach
2. Level of the Standards
3. Form of the Standards
4. Program Flexibilities
5. Coordinated Compliance
C. Summary of Costs and Benefits of the National Program
1. Summary of Costs and Benefits of NHTSA's CAFE Standards
2. Summary of Costs and Benefits of EPA's GHG Standards
D. Background and Comparison of NHTSA and EPA Statutory
Authority
II. Joint Technical Work Completed for This Final Rule

[[Page 25325]]

A. Introduction
B. Developing the Future Fleet for Assessing Costs, Benefits,
and Effects
1. Why did the agencies establish a baseline and reference
vehicle fleet?
2. How did the agencies develop the baseline vehicle fleet?
3. How did the agencies develop the projected MY 2011-2016
vehicle fleet?
4. How was the development of the baseline and reference fleets
for this Final Rule different from NHTSA's historical approach?
5. How does manufacturer product plan data factor into the
baseline used in this Final Rule?
C. Development of Attribute-Based Curve Shapes
D. Relative Car-Truck Stringency
E. Joint Vehicle Technology Assumptions
1. What technologies did the agencies consider?
2. How did the agencies determine the costs and effectiveness of
each of these technologies?
F. Joint Economic Assumptions
G. What are the estimated safety effects of the final MYs 2012-
2016 CAFE and GHG standards?
1. What did the agencies say in the NPRM with regard to
potential safety effects?
2. What public comments did the agencies receive on the safety
analysis and discussions in the NPRM?
3. How has NHTSA refined its analysis for purposes of estimating
the potential safety effects of this Final Rule?
4. What are the estimated safety effects of this Final Rule?
5. How do the agencies plan to address this issue going forward?
III. EPA Greenhouse Gas Vehicle Standards
A. Executive Overview of EPA Rule
1. Introduction
2. Why is EPA establishing this Rule?
3. What is EPA adopting?
4. Basis for the GHG Standards Under Section 202(a)
B. GHG Standards for Light-Duty Vehicles, Light-Duty Trucks, and
Medium-Duty Passenger Vehicles
1. What fleet-wide emissions levels correspond to the
CO2 standards?
2. What are the CO2 attribute-based standards?
3. Overview of How EPA's CO2 Standards Will Be
Implemented for Individual Manufacturers
4. Averaging, Banking, and Trading Provisions for CO2
Standards
5. CO2 Temporary Lead-Time Allowance Alternative
Standards
6. Deferment of CO2 Standards for Small Volume
Manufacturers With Annual Sales Less Than 5,000 Vehicles
7. Nitrous Oxide and Methane Standards
8. Small Entity Exemption
C. Additional Credit Opportunities for CO2 Fleet
Average Program
1. Air Conditioning Related Credits
2. Flexible Fuel and Alternative Fuel Vehicle Credits
3. Advanced Technology Vehicle Incentives for Electric Vehicles,
Plug-in Hybrids, and Fuel Cell Vehicles
4. Off-Cycle Technology Credits
5. Early Credit Options
D. Feasibility of the Final CO2 Standards
1. How did EPA develop a reference vehicle fleet for evaluating
further CO2 reductions?
2. What are the effectiveness and costs of CO2-
reducing technologies?
3. How can technologies be combined into ``packages'' and what
is the cost and effectiveness of packages?
4. Manufacturer's Application of Technology
5. How is EPA projecting that a manufacturer decides between
options to improve CO2 performance to meet a fleet
average standard?
6. Why are the final CO2 standards feasible?
7. What other fleet-wide CO2 levels were considered?
E. Certification, Compliance, and Enforcement
1. Compliance Program Overview
2. Compliance With Fleet-Average CO2 Standards
3. Vehicle Certification
4. Useful Life Compliance
5. Credit Program Implementation
6. Enforcement
7. Prohibited Acts in the CAA
8. Other Certification Issues
9. Miscellaneous Revisions to Existing Regulations
10. Warranty, Defect Reporting, and Other Emission-Related
Components Provisions
11. Light Duty Vehicles and Fuel Economy Labeling
F. How will this Final Rule reduce GHG emissions and their
associated effects?
1. Impact on GHG Emissions
2. Overview of Climate Change Impacts From GHG Emissions
3. Changes in Global Climate Indicators Associated With the
Rule's GHG Emissions Reductions
G. How will the standards impact non-GHG emissions and their
associated effects?
1. Upstream Impacts of Program
2. Downstream Impacts of Program
3. Health Effects of Non-GHG Pollutants
4. Environmental Effects of Non-GHG Pollutants
5. Air Quality Impacts of Non-GHG Pollutants
H. What are the estimated cost, economic, and other impacts of
the program?
1. Conceptual Framework for Evaluating Consumer Impacts
2. Costs Associated With the Vehicle Program
3. Cost per Ton of Emissions Reduced
4. Reduction in Fuel Consumption and Its Impacts
5. Impacts on U.S. Vehicle Sales and Payback Period
6. Benefits of Reducing GHG Emissions
7. Non-Greenhouse Gas Health and Environmental Impacts
8. Energy Security Impacts
9. Other Impacts
10. Summary of Costs and Benefits
I. Statutory and Executive Order Reviews
1. Executive Order 12866: Regulatory Planning and Review
2. Paperwork Reduction Act
3. Regulatory Flexibility Act
4. Unfunded Mandates Reform Act
5. Executive Order 13132 (Federalism)
6. Executive Order 13175 (Consultation and Coordination With
Indian Tribal
Governments)
7. Executive Order 13045: ``Protection of Children From
Environmental Health Risks and Safety Risks''
8. Executive Order 13211 (Energy Effects)
9. National Technology Transfer Advancement Act
10. Executive Order 12898: Federal Actions To Address
Environmental Justice in Minority Populations and Low-Income
Populations
J. Statutory Provisions and Legal Authority
IV. NHTSA Final Rule and Record of Decision for Passenger Car and
Light Truck CAFE Standards for MYs 2012-2016
A. Executive Overview of NHTSA Final Rule
1. Introduction
2. Role of Fuel Economy Improvements in Promoting Energy
Independence, Energy Security, and a Low Carbon Economy
3. The National Program
4. Review of CAFE Standard Setting Methodology per the
President's January 26, 2009 Memorandum on CAFE Standards for MYs
2011 and Beyond
5. Summary of the Final MY 2012-2016 CAFE Standards
B. Background
1. Chronology of Events Since the National Academy of Sciences
Called for Reforming and Increasing CAFE Standards
2. Energy Policy and Conservation Act, as Amended by the Energy
Independence and Security Act
C. Development and Feasibility of the Final Standards
1. How was the baseline and reference vehicle fleet developed?
2. How were the technology inputs developed?
3. How did NHTSA develop the economic assumptions?
4. How does NHTSA use the assumptions in its modeling analysis?
5. How did NHTSA develop the shape of the target curves for the
final standards?
D. Statutory Requirements
1. EPCA, as Amended by EISA
2. Administrative Procedure Act
3. National Environmental Policy Act
E. What are the final CAFE standards?
1. Form of the Standards
2. Passenger Car Standards for MYs 2012-2016
3. Minimum Domestic Passenger Car Standards
4. Light Truck Standards
F. How do the final standards fulfill NHTSA's statutory
obligations?
G. Impacts of the Final CAFE Standards
1. How will these standards improve fuel economy and reduce GHG
emissions for MY 2012-2016 vehicles?
2. How will these standards improve fleet-wide fuel economy and
reduce GHG emissions beyond MY 2016?

[[Page 25326]]

3. How will these final standards impact non-GHG emissions and
their associated effects?
4. What are the estimated costs and benefits of these final
standards?
5. How would these standards impact vehicle sales?
6. Potential Unquantified Consumer Welfare Impacts of the Final
Standards
7. What other impacts (quantitative and unquantifiable) will
these final standards have?
H. Vehicle Classification
I. Compliance and Enforcement
1. Overview
2. How does NHTSA determine compliance?
3. What compliance flexibilities are available under the CAFE
program and how do manufacturers use them?
4. Other CAFE Enforcement Issues--Variations in Footprint
5. Other CAFE Enforcement Issues--Miscellaneous
J. Other Near-Term Rulemakings Mandated by EISA
1. Commercial Medium- and Heavy-Duty On-Highway Vehicles and
Work Trucks
2. Consumer Information on Fuel Efficiency and Emissions
K. NHTSA's Record of Decision
L. Regulatory Notices and Analyses
1. Executive Order 12866 and DOT Regulatory Policies and
Procedures
2. National Environmental Policy Act
3. Clean Air Act (CAA)
4. National Historic Preservation Act (NHPA)
5. Executive Order 12898 (Environmental Justice)
6. Fish and Wildlife Conservation Act (FWCA)
7. Coastal Zone Management Act (CZMA)
8. Endangered Species Act (ESA)
9. Floodplain Management (Executive Order 11988 & DOT Order
5650.2)
10. Preservation of the Nation's Wetlands (Executive Order 11990
& DOT Order 5660.1a)
11. Migratory Bird Treaty Act (MBTA), Bald and Golden Eagle
Protection Act (BGEPA), Executive Order 13186
12. Department of Transportation Act (Section 4(f))
13. Regulatory Flexibility Act
14. Executive Order 13132 (Federalism)
15. Executive Order 12988 (Civil Justice Reform)
16. Unfunded Mandates Reform Act
17. Regulation Identifier Number
18. Executive Order 13045
19. National Technology Transfer and Advancement Act
20. Executive Order 13211
21. Department of Energy Review
22. Privacy Act

I. Overview of Joint EPA/NHTSA National Program

A. Introduction

The National Highway Traffic Safety Administration (NHTSA) and the
Environmental Protection Agency (EPA) are each announcing final rules
whose benefits will address the urgent and closely intertwined
challenges of energy independence and security and global warming.
These rules will implement a strong and coordinated Federal greenhouse
gas (GHG) and fuel economy program for passenger cars, light-duty-
trucks, and medium-duty passenger vehicles (hereafter light-duty
vehicles), referred to as the National Program. The rules will achieve
substantial reductions of GHG emissions and improvements in fuel
economy from the light-duty vehicle part of the transportation sector,
based on technology that is already being commercially applied in most
cases and that can be incorporated at a reasonable cost. NHTSA's final
rule also constitutes the agency's Record of Decision for purposes of
its NEPA analysis.
This joint rulemaking is consistent with the President's
announcement on May 19, 2009 of a National Fuel Efficiency Policy of
establishing consistent, harmonized, and streamlined requirements that
would reduce GHG emissions and improve fuel economy for all new cars
and light-duty trucks sold in the United States.\3\ The National
Program will deliver additional environmental and energy benefits, cost
savings, and administrative efficiencies on a nationwide basis that
would likely not be available under a less coordinated approach. The
National Program also represents regulatory convergence by making it
possible for the standards of two different Federal agencies and the
standards of California and other states to act in a unified fashion in
providing these benefits. The National Program will allow automakers to
produce and sell a single fleet nationally, mitigating the additional
costs that manufacturers would otherwise face in having to comply with
multiple sets of Federal and State standards. This joint notice is also
consistent with the Notice of Upcoming Joint Rulemaking issued by DOT
and EPA on May 19, 2009 \4\ and responds to the President's January 26,
2009 memorandum on CAFE standards for model years 2011 and beyond,\5\
the details of which can be found in Section IV of this joint notice.
---------------------------------------------------------------------------

\3\ President Obama Announces National Fuel Efficiency Policy,
The White House, May 19, 2009. Available at: http://
www.whitehouse.gov/the_press_office/President-Obama-Announces-
National-Fuel-Efficiency-Policy/. Remarks by the President on
National Fuel Efficiency Standards, The White House, May 19, 2009.
Available at: http://www.whitehouse.gov/the_press_office/Remarks-
by-the-President-on-national-fuel-efficiency-standards/.
\4\ 74 FR 24007 (May 22, 2009).
\5\ Available at: http://www.whitehouse.gov/the_press_office/
Presidential_Memorandum_Fuel_Economy/.
---------------------------------------------------------------------------

Climate change is widely viewed as a significant long-term threat
to the global environment. As summarized in the Technical Support
Document for EPA's Endangerment and Cause or Contribute Findings under
Section 202(a) of the Clear Air Act, anthropogenic emissions of GHGs
are very likely (90 to 99 percent probability) the cause of most of the
observed global warming over the last 50 years.\6\ The primary GHGs of
concern are carbon dioxide (CO2), methane, nitrous oxide,
hydrofluorocarbons, perfluorocarbons, and sulfur hexafluoride. Mobile
sources emitted 31 percent of all U.S. GHGs in 2007 (transportation
sources, which do not include certain off-highway sources, account for
28 percent) and have been the fastest-growing source of U.S. GHGs since
1990.\7\ Mobile sources addressed in the recent endangerment and
contribution findings under CAA section 202(a)--light-duty vehicles,
heavy-duty trucks, buses, and motorcycles--accounted for 23 percent of
all U.S. GHG in 2007.\8\ Light-duty vehicles emit CO2,
methane, nitrous oxide, and hydrofluorocarbons and are responsible for
nearly 60 percent of all mobile source GHGs and over 70 percent of
Section 202(a) mobile source GHGs. For light-duty vehicles in 2007,
CO2 emissions represent about 94 percent of all greenhouse
emissions (including HFCs), and the CO2 emissions measured
over the EPA tests used for fuel economy compliance represent about 90
percent of total light-duty vehicle GHG emissions.^{9 10}
---------------------------------------------------------------------------

\6\ ``Technical Support Document for Endangerment and Cause or
Contribute Findings for Greenhouse Gases Under Section 202(a) of the
Clean Air Act'' Docket: EPA-HQ-OAR-2009-0472-11292, http://epa.gov/
climatechange/endangerment.html.
\7\ U.S. Environmental Protection Agency. 2009. Inventory of
U.S. Greenhouse Gas Emissions and Sinks: 1990-2007. EPA 430-R-09-
004. Available at http://epa.gov/climatechange/emissions/
downloads09/GHG2007entire_report-508.pdf.
\8\ U.S. EPA. 2009 Technical Support Document for Endangerment
and Cause or Contribute Findings for Greenhouse Gases under Section
202(a) of the Clean Air Act. Washington, DC. pp. 180-194. Available
at http://epa.gov/climatechange/endangerment/downloads/
Endangerment%20TSD.pdf.
\9\ U.S. Environmental Protection Agency. 2009. Inventory of
U.S. Greenhouse Gas Emissions and Sinks: 1990-2007. EPA 430-R-09-
004. Available at http://epa.gov/climatechange/emissions/
downloads09/GHG2007entire_report-508.pdf.
\10\ U.S. Environmental Protection Agency. RIA, Chapter 2.
---------------------------------------------------------------------------

Improving energy security by reducing our dependence on foreign oil
has been a national objective since the first oil price shocks in the
1970s. Net petroleum imports now account for approximately 60 percent
of U.S.

[[Page 25327]]

petroleum consumption. World crude oil production is highly
concentrated, exacerbating the risks of supply disruptions and price
shocks. Tight global oil markets led to prices over $100 per barrel in
2008, with gasoline reaching as high as $4 per gallon in many parts of
the U.S., causing financial hardship for many families. The export of
U.S. assets for oil imports continues to be an important component of
the historically unprecedented U.S. trade deficits. Transportation
accounts for about two-thirds of U.S. petroleum consumption. Light-duty
vehicles account for about 60 percent of transportation oil use, which
means that they alone account for about 40 percent of all U.S. oil
consumption.
1. Building Blocks of the National Program
The National Program is both needed and possible because the
relationship between improving fuel economy and reducing CO2
tailpipe emissions is a very direct and close one. The amount of those
CO2 emissions is essentially constant per gallon combusted
of a given type of fuel. Thus, the more fuel efficient a vehicle is,
the less fuel it burns to travel a given distance. The less fuel it
burns, the less CO2 it emits in traveling that distance.\11\
While there are emission control technologies that reduce the
pollutants (e.g., carbon monoxide) produced by imperfect combustion of
fuel by capturing or converting them to other compounds, there is no
such technology for CO2. Further, while some of those
pollutants can also be reduced by achieving a more complete combustion
of fuel, doing so only increases the tailpipe emissions of
CO2. Thus, there is a single pool of technologies for
addressing these twin problems, i.e., those that reduce fuel
consumption and thereby reduce CO2 emissions as well.
---------------------------------------------------------------------------

\11\ Panel on Policy Implications of Greenhouse Warming,
National Academy of Sciences, National Academy of Engineering,
Institute of Medicine, ``Policy Implications of Greenhouse Warming:
Mitigation, Adaptation, and the Science Base,'' National Academies
Press, 1992. p. 287.
---------------------------------------------------------------------------

a. DOT's CAFE Program
In 1975, Congress enacted the Energy Policy and Conservation Act
(EPCA), mandating that NHTSA establish and implement a regulatory
program for motor vehicle fuel economy to meet the various facets of
the need to conserve energy, including ones having energy independence
and security, environmental and foreign policy implications. Fuel
economy gains since 1975, due both to the standards and market factors,
have resulted in saving billions of barrels of oil and avoiding
billions of metric tons of CO2 emissions. In December 2007,
Congress enacted the Energy Independence and Securities Act (EISA),
amending EPCA to require substantial, continuing increases in fuel
economy standards.
The CAFE standards address most, but not all, of the real world
CO2 emissions because a provision in EPCA as originally
enacted in 1975 requires the use of the 1975 passenger car test
procedures under which vehicle air conditioners are not turned on
during fuel economy testing.\12\ Fuel economy is determined by
measuring the amount of CO2 and other carbon compounds
emitted from the tailpipe, not by attempting to measure directly the
amount of fuel consumed during a vehicle test, a difficult task to
accomplish with precision. The carbon content of the test fuel \13\ is
then used to calculate the amount of fuel that had to be consumed per
mile in order to produce that amount of CO2. Finally, that
fuel consumption figure is converted into a miles-per-gallon figure.
CAFE standards also do not address the 5-8 percent of GHG emissions
that are not CO2, i.e., nitrous oxide (N2O), and
methane (CH4) as well as emissions of CO2 and
hydrofluorocarbons (HFCs) related to operation of the air conditioning
system.
---------------------------------------------------------------------------

\12\ Although EPCA does not require the use of 1975 test
procedures for light trucks, those procedures are used for light
truck CAFE standard testing purposes.
\13\ This is the method that EPA uses to determine compliance
with NHTSA's CAFE standards.
---------------------------------------------------------------------------

b. EPA's GHG Standards for Light-duty Vehicles
Under the Clean Air Act EPA is responsible for addressing air
pollutants from motor vehicles. On April 2, 2007, the U.S. Supreme
Court issued its opinion in Massachusetts v. EPA,\14\ a case involving
EPA's a 2003 denial of a petition for rulemaking to regulate GHG
emissions from motor vehicles under section 202(a) of the Clean Air Act
(CAA).\15\ The Court held that GHGs fit within the definition of air
pollutant in the Clean Air Act and further held that the Administrator
must determine whether or not emissions from new motor vehicles cause
or contribute to air pollution which may reasonably be anticipated to
endanger public health or welfare, or whether the science is too
uncertain to make a reasoned decision. The Court further ruled that, in
making these decisions, the EPA Administrator is required to follow the
language of section 202(a) of the CAA. The Court rejected the argument
that EPA cannot regulate CO2 from motor vehicles because to
do so would de facto tighten fuel economy standards, authority over
which has been assigned by Congress to DOT. The Court stated that
``[b]ut that DOT sets mileage standards in no way licenses EPA to shirk
its environmental responsibilities. EPA has been charged with
protecting the public's `health' and `welfare', a statutory obligation
wholly independent of DOT's mandate to promote energy efficiency.'' The
Court concluded that ``[t]he two obligations may overlap, but there is
no reason to think the two agencies cannot both administer their
obligations and yet avoid inconsistency.'' \16\ The case was remanded
back to the Agency for reconsideration in light of the Court's
decision.\17\
---------------------------------------------------------------------------

\14\ 549 U.S. 497 (2007).
\15\ 68 FR 52922 (Sept. 8, 2003).
\16\ 549 U.S. at 531-32.
\17\ For further information on Massachusetts v. EPA see the
July 30, 2008 Advance Notice of Proposed Rulemaking, ``Regulating
Greenhouse Gas Emissions under the Clean Air Act'', 73 FR 44354 at
44397. There is a comprehensive discussion of the litigation's
history, the Supreme Court's findings, and subsequent actions
undertaken by the Bush Administration and the EPA from 2007-2008 in
response to the Supreme Court remand. Also see 74 FR 18886, at 1888-
90 (April 24, 2009).
---------------------------------------------------------------------------

On December 15, 2009, EPA published two findings (74 FR 66496):
That emissions of GHGs from new motor vehicles and motor vehicle
engines contribute to air pollution, and that the air pollution may
reasonably be anticipated to endanger public health and welfare.
c. California Air Resources Board Greenhouse Gas Program
In 2004, the California Air Resources Board approved standards for
new light-duty vehicles, which regulate the emission of not only
CO2, but also other GHGs. Since then, thirteen states and
the District of Columbia, comprising approximately 40 percent of the
light-duty vehicle market, have adopted California's standards. These
standards apply to model years 2009 through 2016 and require
CO2 emissions for passenger cars and the smallest light
trucks of 323 g/mi in 2009 and 205 g/mi in 2016, and for the remaining
light trucks of 439 g/mi in 2009 and 332 g/mi in 2016. On June 30,
2009, EPA granted California's request for a waiver of preemption under
the CAA.\18\ The granting of the waiver permits California and the
other states to proceed with implementing the California emission
standards.
---------------------------------------------------------------------------

\18\ 74 FR 32744 (July 8, 2009).
---------------------------------------------------------------------------

In addition, to promote the National Program, in May 2009,
California announced its commitment to take several actions in support
of the National Program, including revising its

[[Page 25328]]

program for MYs 2009-2011 to facilitate compliance by the automakers,
and revising its program for MYs 2012-2016 such that compliance with
the Federal GHG standards will be deemed to be compliance with
California's GHG standards. This will allow the single national fleet
produced by automakers to meet the two Federal requirements and to meet
California requirements as well. California is proceeding with a
rulemaking intended to revise its 2004 regulations to meet its
commitments. Several automakers and their trade associations also
announced their commitment to take several actions in support of the
National Program, including not contesting the final GHG and CAFE
standards for MYs 2012-2016, not contesting any grant of a waiver of
preemption under the CAA for California's GHG standards for certain
model years, and to stay and then dismiss all pending litigation
challenging California's regulation of GHG emissions, including
litigation concerning preemption under EPCA of California's and other
states' GHG standards.
2. Public Participation
The agencies proposed their respective rules on September 28, 2009
(74 FR 49454), and received a large number of comments representing
many perspectives on the proposed rule. The agencies received oral
testimony at three public hearings in different parts of the country,
and received written comments from more than 130 organizations,
including auto manufacturers and suppliers, States, environmental and
other non-governmental organizations (NGOs), and over 129,000 comments
from private citizens.
The vast majority of commenters supported the central tenets of the
proposed CAFE and GHG programs. That is, there was broad support from
most organizations for a National Program that achieves a level of 250
gram/mile fleet average CO2, which would be 35.5 miles per
gallon if the automakers were to meet this CO2 level solely
through fuel economy improvements. The standards will be phased in over
model years 2012 through 2016 which will allow manufacturers to build a
common fleet of vehicles for the domestic market. In general,
commenters from the automobile industry supported the proposed
standards as well as the credit opportunities and other compliance
provisions providing flexibility, while also making some
recommendations for changes. Environmental and public interest non-
governmental organizations (NGOs), as well as most States that
commented, were also generally supportive of the National Program
standards. Many of these organizations also expressed concern about the
possible impact on program benefits, depending on how the credit
provisions and flexibilities are designed. The agencies also received
specific comments on many aspects of the proposal.
Throughout this notice, the agencies discuss many of the key issues
arising from the public comments and the agencies' responses. In
addition, the agencies have addressed all of the public comments in the
Response to Comments document associated with this final rule.

B. Summary of the Joint Final Rule and Differences From the Proposal

In this joint rulemaking, EPA is establishing GHG emissions
standards under the Clean Air Act (CAA), and NHTSA is establishing
Corporate Average Fuel Economy (CAFE) standards under the Energy Policy
and Conservation Action of 1975 (EPCA), as amended by the Energy
Independence and Security Act of 2007 (EISA). The intention of this
joint rulemaking is to set forth a carefully coordinated and harmonized
approach to implementing these two statutes, in accordance with all
substantive and procedural requirements imposed by law.
NHTSA and EPA have coordinated closely and worked jointly in
developing their respective final rules. This is reflected in many
aspects of this joint rule. For example, the agencies have developed a
comprehensive Joint Technical Support Document (TSD) that provides a
solid technical underpinning for each agency's modeling and analysis
used to support their standards. Also, to the extent allowed by law,
the agencies have harmonized many elements of program design, such as
the form of the standard (the footprint-based attribute curves), and
the definitions used for cars and trucks. They have developed the same
or similar compliance flexibilities, to the extent allowed and
appropriate under their respective statutes, such as averaging,
banking, and trading of credits, and have harmonized the compliance
testing and test protocols used for purposes of the fleet average
standards each agency is finalizing. Finally, under their respective
statutes, each agency is called upon to exercise its judgment and
determine standards that are an appropriate balance of various relevant
statutory factors. Given the common technical issues before each
agency, the similarity of the factors each agency is to consider and
balance, and the authority of each agency to take into consideration
the standards of the other agency, both EPA and NHTSA are establishing
standards that result in a harmonized National Program.
This joint final rule covers passenger cars, light-duty trucks, and
medium-duty passenger vehicles built in model years 2012 through 2016.
These vehicle categories are responsible for almost 60 percent of all
U.S. transportation-related GHG emissions. EPA and NHTSA expect that
automobile manufacturers will meet these standards by utilizing
technologies that will reduce vehicle GHG emissions and improve fuel
economy. Although many of these technologies are available today, the
emissions reductions and fuel economy improvements finalized in this
notice will involve more widespread use of these technologies across
the light-duty vehicle fleet. These include improvements to engines,
transmissions, and tires, increased use of start-stop technology,
improvements in air conditioning systems, increased use of hybrid and
other advanced technologies, and the initial commercialization of
electric vehicles and plug-in hybrids. NHTSA's and EPA's assessments of
likely vehicle technologies that manufacturers will employ to meet the
standards are discussed in detail below and in the Joint TSD.
The National Program is estimated to result in approximately 960
million metric tons of total carbon dioxide equivalent emissions
reductions and approximately 1.8 billion barrels of oil savings over
the lifetime of vehicles sold in model years (MYs) 2012 through 2016.
In total, the combined EPA and NHTSA 2012-2016 standards will reduce
GHG emissions from the U.S. light-duty fleet by approximately 21
percent by 2030 over the level that would occur in the absence of the
National Program. These actions also will provide important energy
security benefits, as light-duty vehicles are about 95 percent
dependent on oil-based fuels. The agencies project that the total
benefits of the National Program will be more than $240 billion at a 3%
discount rate, or more than $190 billion at a 7% discount rate. In the
discussion that follows in Sections III and IV, each agency explains
the related benefits for their individual standards.
Together, EPA and NHTSA estimate that the average cost increase for
a model year 2016 vehicle due to the National Program will be less than
$1,000. The average U.S. consumer who purchases a vehicle outright is
estimated to save enough in lower fuel costs over the first three years
to offset

[[Page 25329]]

these higher vehicle costs. However, most U.S. consumers purchase a new
vehicle using credit rather than paying cash and the typical car loan
today is a five year, 60 month loan. These consumers will see immediate
savings due to their vehicle's lower fuel consumption in the form of a
net reduction in annual costs of $130-$180 throughout the duration of
the loan (that is, the fuel savings will outweigh the increase in loan
payments by $130-$180 per year). Whether a consumer takes out a loan or
purchases a new vehicle outright, over the lifetime of a model year
2016 vehicle, the consumer's net savings could be more than $3,000. The
average 2016 MY vehicle will emit 16 fewer metric tons of
CO2-equivalent emissions (that is, CO2 emissions
plus HFC air conditioning leakage emissions) during its lifetime.
Assumptions that underlie these conclusions are discussed in greater
detail in the agencies' respective regulatory impact analyses and in
Section III.H.5 and Section IV.
This joint rule also results in important regulatory convergence
and certainty to automobile companies. Absent this rule, there would be
three separate Federal and State regimes independently regulating
light-duty vehicles to reduce fuel consumption and GHG emissions:
NHTSA's CAFE standards, EPA's GHG standards, and the GHG standards
applicable in California and other States adopting the California
standards. This joint rule will allow automakers to meet both the NHTSA
and EPA requirements with a single national fleet, greatly simplifying
the industry's technology, investment and compliance strategies. In
addition, to promote the National Program, California announced its
commitment to take several actions, including revising its program for
MYs 2012-2016 such that compliance with the Federal GHG standards will
be deemed to be compliance with California's GHG standards. This will
allow the single national fleet used by automakers to meet the two
Federal requirements and to meet California requirements as well.
California is proceeding with a rulemaking intended to revise its 2004
regulations to meet its commitments. EPA and NHTSA are confident that
these GHG and CAFE standards will successfully harmonize both the
Federal and State programs for MYs 2012-2016 and will allow our country
to achieve the increased benefits of a single, nationwide program to
reduce light-duty vehicle GHG emissions and reduce the country's
dependence on fossil fuels by improving these vehicles' fuel economy.
A successful and sustainable automotive industry depends upon,
among other things, continuous technology innovation in general, and
low GHG emissions and high fuel economy vehicles in particular. In this
respect, this action will help spark the investment in technology
innovation necessary for automakers to successfully compete in both
domestic and export markets, and thereby continue to support a strong
economy.
While this action covers MYs 2012-2016, many stakeholders
encouraged EPA and NHTSA to also begin working toward standards for MY
2017 and beyond that would maintain a single nationwide program. The
agencies recognize the importance of and are committed to a strong,
coordinated national program for light-duty vehicles for model years
beyond 2016.
Key elements of the National Program finalized today are the level
and form of the GHG and CAFE standards, the available compliance
mechanisms, and general implementation elements. These elements are
summarized in the following section, with more detailed discussions
about EPA's GHG program following in Section III, and about NHTSA's
CAFE program in Section IV. This joint final rule responds to the wide
array of comments that the agencies received on the proposed rule. This
section summarizes many of the major comments on the primary elements
of the proposal and describes whether and how the final rule has
changed, based on the comments and additional analyses. Major comments
and the agencies' responses to them are also discussed in more detail
in later sections of this preamble. For a full summary of public
comments and EPA's and NHTSA's responses to them, please see the
Response to Comments document associated with this final rule.
1. Joint Analytical Approach
NHTSA and EPA have worked closely together on nearly every aspect
of this joint final rule. The extent and results of this collaboration
are reflected in the elements of the respective NHTSA and EPA rules, as
well as the analytical work contained in the Joint Technical Support
Document (Joint TSD). The Joint TSD, in particular, describes important
details of the analytical work that are shared, as well as any
differences in approach. These include the build up of the baseline and
reference fleets, the derivation of the shape of the curves that define
the standards, a detailed description of the costs and effectiveness of
the technology choices that are available to vehicle manufacturers, a
summary of the computer models used to estimate how technologies might
be added to vehicles, and finally the economic inputs used to calculate
the impacts and benefits of the rules, where practicable.
EPA and NHTSA have jointly developed attribute curve shapes that
each agency is using for its final standards. Further details of these
functions can be found in Sections III and IV of this preamble as well
as Chapter 2 of the Joint TSD. A critical technical underpinning of
each agency's analysis is the cost and effectiveness of the various
control technologies. These are used to analyze the feasibility and
cost of potential GHG and CAFE standards. A detailed description of all
of the technology information considered can be found in Chapter 3 of
the Joint TSD (and for A/C, Chapter 2 of the EPA RIA). This detailed
technology data forms the inputs to computer models that each agency
uses to project how vehicle manufacturers may add those technologies in
order to comply with the new standards. These are the OMEGA and Volpe
models for EPA and NHTSA, respectively. The models and their inputs can
also be found in the docket. Further description of the model and
outputs can be found in Sections III and IV of this preamble, and
Chapter 3 of the Joint TSD. This comprehensive joint analytical
approach has provided a sound and consistent technical basis for each
agency in developing its final standards, which are summarized in the
sections below.
The vast majority of public comments expressed strong support for
the joint analytical work performed for the proposal. Commenters
generally agreed with the analytical work and its results, and
supported the transparency of the analysis and its underlying data.
Where commenters raised specific points, the agencies have considered
them and made changes where appropriate. The agencies' further
evaluation of various technical issues also led to a limited number of
changes. A detailed discussion of these issues can be found in Section
II of this preamble, and the Joint TSD.
2. Level of the Standards
In this notice, EPA and NHTSA are establishing two separate sets of
standards, each under its respective statutory authorities. EPA is
setting national CO2 emissions standards for light-duty
vehicles under section 202(a) of the Clean Air Act. These standards
will require these vehicles to meet an

[[Page 25330]]

estimated combined average emissions level of 250 grams/mile of
CO2 in model year 2016. NHTSA is setting CAFE standards for
passenger cars and light trucks under 49 U.S.C. 32902. These standards
will require manufacturers of those vehicles to meet an estimated
combined average fuel economy level of 34.1 mpg in model year 2016. The
standards for both agencies begin with the 2012 model year, with
standards increasing in stringency through model year 2016. They
represent a harmonized approach that will allow industry to build a
single national fleet that will satisfy both the GHG requirements under
the CAA and CAFE requirements under EPCA/EISA.
Given differences in their respective statutory authorities,
however, the agencies' standards include some important differences.
Under the CO2 fleet average standards adopted under CAA
section 202(a), EPA expects manufacturers to take advantage of the
option to generate CO2-equivalent credits by reducing
emissions of hydrofluorocarbons (HFCs) and CO2 through
improvements in their air conditioner systems. EPA accounted for these
reductions in developing its final CO2 standards. NHTSA did
not do so because EPCA does not allow vehicle manufacturers to use air
conditioning credits in complying with CAFE standards for passenger
cars.\19\ CO2 emissions due to air conditioning operation
are not measured by the test procedure mandated by statute for use in
establishing and enforcing CAFE standards for passenger cars. As a
result, improvement in the efficiency of passenger car air conditioners
is not considered as a possible control technology for purposes of
CAFE.
---------------------------------------------------------------------------

\19\ There is no such statutory limitation with respect to light
trucks.
---------------------------------------------------------------------------

These differences regarding the treatment of air conditioning
improvements (related to CO2 and HFC reductions) affect the
relative stringency of the EPA standard and NHTSA standard for MY 2016.
The 250 grams per mile of CO2 equivalent emissions limit is
equivalent to 35.5 mpg \20\ if the automotive industry were to meet
this CO2 level all through fuel economy improvements. As a
consequence of the prohibition against NHTSA's allowing credits for air
conditioning improvements for purposes of passenger car CAFE
compliance, NHTSA is setting fuel economy standards that are estimated
to require a combined (passenger car and light truck) average fuel
economy level of 34.1 mpg by MY 2016.
---------------------------------------------------------------------------

\20\ The agencies are using a common conversion factor between
fuel economy in units of miles per gallon and CO2
emissions in units of grams per mile. This conversion factor is
8,887 grams CO2 per gallon gasoline fuel. Diesel fuel has
a conversion factor of 10,180 grams CO2 per gallon diesel
fuel though for the purposes of this calculation, we are assuming
100% gasoline fuel.
---------------------------------------------------------------------------

The vast majority of public comments expressed strong support for
the National Program standards, including the stringency of the
agencies' respective standards and the phase-in from model year 2012
through 2016. There were a number of comments supporting standards more
stringent than proposed, and a few others supporting less stringent
standards, in particular for the 2012-2015 model years. The agencies'
consideration of comments and their updated technical analyses led to
only very limited changes in the footprint curves and did not change
the agencies' projections that the nationwide fleet will achieve a
level of 250 grams/mile by 2016 (equivalent to 35.5 mpg). The responses
to these comments are discussed in more detail in Sections III and IV,
respectively, and in the Response to Comments document.
As proposed, NHTSA and EPA's final standards, like the standards
NHTSA promulgated in March 2009 for MY 2011, are expressed as
mathematical functions depending on vehicle footprint. Footprint is one
measure of vehicle size, and is determined by multiplying the vehicle's
wheelbase by the vehicle's average track width.\21\ The standards that
must be met by each manufacturer's fleet will be determined by
computing the sales-weighted average (harmonic average for CAFE) of the
targets applicable to each of the manufacturer's passenger cars and
light trucks. Under these footprint-based standards, the levels
required of individual manufacturers will depend, as noted above, on
the mix of vehicles sold. NHTSA's and EPA's respective standards are
shown in the tables below. It is important to note that the standards
are the attribute-based curves established by each agency. The values
in the tables below reflect the agencies' projection of the
corresponding fleet levels that will result from these attribute-based
curves.
---------------------------------------------------------------------------

\21\ See 49 CFR 523.2 for the exact definition of ``footprint.''
---------------------------------------------------------------------------

As a result of public comments and updated economic and future
fleet projections, EPA and NHTSA have updated the attribute based
curves for this final rule, as discussed in detail in Section II.B of
this preamble and Chapter 2 of the Joint TSD. This update in turn
affects costs, benefits, and other impacts of the final standards.
Thus, the agencies have updated their overall projections of the
impacts of the final rule standards, and these results are only
slightly different from those presented in the proposed rule.
As shown in Table I.B.2-1, NHTSA's fleet-wide CAFE-required levels
for passenger cars under the final standards are projected to increase
from 33.3 to 37.8 mpg between MY 2012 and MY 2016. Similarly, fleet-
wide CAFE levels for light trucks are projected to increase from 25.4
to 28.8 mpg. NHTSA has also estimated the average fleet-wide required
levels for the combined car and truck fleets. As shown, the overall
fleet average CAFE level is expected to be 34.1 mpg in MY 2016. These
numbers do not include the effects of other flexibilities and credits
in the program. These standards represent a 4.3 percent average annual
rate of increase relative to the MY 2011 standards.\22\
---------------------------------------------------------------------------

\22\ Because required CAFE levels depend on the mix of vehicles
sold by manufacturers in a model year, NHTSA's estimate of future
required CAFE levels depends on its estimate of the mix of vehicles
that will be sold in that model year. NHTSA currently estimates that
the MY 2011 standards will require average fuel economy levels of
30.4 mpg for passenger cars, 24.4 mpg for light trucks, and 27.6 mpg
for the combined fleet.

Table I.B.2-1--Average Required Fuel Economy (mpg) Under Final CAFE Standards
--------------------------------------------------------------------------------------------------------------------------------------------------------
2011-base 2012 2013 2014 2015 2016
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 30.4 33.3 34.2 34.9 36.2 37.8
Light Trucks............................................ 24.4 25.4 26.0 26.6 27.5 28.8
-----------------------------------------------------------------------------------------------
Combined Cars & Trucks.............................. 27.6 29.7 30.5 31.3 32.6 34.1
--------------------------------------------------------------------------------------------------------------------------------------------------------

[[Page 25331]]

Accounting for the expectation that some manufacturers could
continue to pay civil penalties rather than achieving required CAFE
levels, and the ability to use FFV credits,\23\ NHTSA estimates that
the CAFE standards will lead to the following average achieved fuel
economy levels, based on the projections of what each manufacturer's
fleet will comprise in each year of the program: \24\
---------------------------------------------------------------------------

\23\ The penalties are similar in function to essentially
unlimited, fixed-price allowances.
\24\ NHTSA's estimates account for availability of CAFE credits
for the sale of flexible-fuel vehicles (FFVs), and for the potential
that some manufacturers will pay civil penalties rather than comply
with the CAFE standards. This yields NHTSA's estimates of the real-
world fuel economy that will likely be achieved under the final CAFE
standards. NHTSA has not included any potential impact of car-truck
credit transfer in its estimate of the achieved CAFE levels.

Table I.B.2-2--Projected Fleet-Wide Achieved CAFE Levels Under the Final Footprint-Based CAFE Standards (mpg)
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................. 32.3 33.5 34.2 35.0 36.2
Light Trucks.................... 24.5 25.1 25.9 26.7 27.5
-------------------------------------------------------------------------------
Combined Cars & Trucks...... 28.7 29.7 30.6 31.5 32.7
----------------------------------------------------------------------------------------------------------------

NHTSA is also required by EISA to set a minimum fuel economy
standard for domestically manufactured passenger cars in addition to
the attribute-based passenger car standard. The minimum standard
``shall be the greater of (A) 27.5 miles per gallon; or (B) 92 percent
of the average fuel economy projected by the Secretary for the combined
domestic and non-domestic passenger automobile fleets manufactured for
sale in the United States by all manufacturers in the model year.* * *
'' \25\
---------------------------------------------------------------------------

\25\ 49 U.S.C. 32902(b)(4).
---------------------------------------------------------------------------

Based on NHTSA's current market forecast, the agency's estimates of
these minimum standards under the MY 2012-2016 CAFE standards (and, for
comparison, the final MY 2011 standard) are summarized below in Table
I.B.2-3.\26\ For eventual compliance calculations, the final calculated
minimum standards will be updated to reflect the average fuel economy
level required under the final standards.
---------------------------------------------------------------------------

\26\ In the March 2009 final rule establishing MY 2011 standards
for passenger cars and light trucks, NHTSA estimated that the
minimum required CAFE standard for domestically manufactured
passenger cars would be 27.8 mpg under the MY 2011 passenger car
standard.

Table I.B.2-3--Estimated Minimum Standard for Domestically Manufactured Passenger Cars Under MY 2011 and MY 2012-
2016 CAFE Standards for Passenger Cars (mpg)
----------------------------------------------------------------------------------------------------------------
2011 2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
27.8 30.7 31.4 32.1 33.3 34.7
----------------------------------------------------------------------------------------------------------------

EPA is establishing GHG emissions standards, and Table I.B.2-4
provides EPA's estimates of their projected overall fleet-wide
CO2 equivalent emission levels.\27\ The g/mi values are
CO2 equivalent values because they include the projected use
of air conditioning (A/C) credits by manufacturers, which include both
HFC and CO2 reductions.
---------------------------------------------------------------------------

\27\ These levels do not include the effect of flexible fuel
credits, transfer of credits between cars and trucks, temporary lead
time allowance, or any other credits with the exception of air
conditioning.

Table I.B.2-4--Projected Fleet-Wide Emissions Compliance Levels Under the Footprint-Based CO2 Standards (g/mi)
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................. 263 256 247 236 225
Light Trucks.................... 346 337 326 312 298
-------------------------------------------------------------------------------
Combined Cars & Trucks...... 295 286 276 263 250
----------------------------------------------------------------------------------------------------------------

As shown in Table I.B.2-4, fleet-wide CO2 emission level
requirements for cars are projected to increase in stringency from 263
to 225 g/mi between MY 2012 and MY 2016. Similarly, fleet-wide
CO2 equivalent emission level requirements for trucks are
projected to increase in stringency from 346 to 298 g/mi. As shown, the
overall fleet average CO2 level requirements are projected
to increase in stringency from 295 g/mi in MY 2012 to 250 g/mi in MY
2016.
EPA anticipates that manufacturers will take advantage of program
flexibilities such as flexible fueled vehicle credits and car/truck
credit trading. Due to the credit trading between cars and trucks, the
estimated improvements in CO2 emissions are distributed
differently than shown in Table I.B.2-4, where full manufacturer
compliance without credit trading is assumed. Table I.B.2-5 shows EPA's
projection of the achieved emission levels of the fleet for MY 2012
through 2016, which does consider the impact of car/truck credit
transfer and the increase in emissions due to certain program
flexibilities including flex fueled vehicle credits and the temporary
lead time allowance alternative standards. The use of optional air
conditioning credits is considered both in this analysis of achieved
levels and of the

[[Page 25332]]

compliance levels described above. As can be seen in Table I.B.2-5, the
projected achieved levels are slightly higher for model years 2012-2015
due to EPA's assumptions about manufacturers' use of the regulatory
flexibilities, but by model year 2016 the achieved level is projected
to be 250 g/mi for the fleet.

Table I.B.2-5--Projected Fleet-Wide Achieved Emission Levels Under the Footprint-Based CO2 Standards (g/mi)
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................. 267 256 245 234 223
Light Trucks.................... 365 353 340 324 303
-------------------------------------------------------------------------------
Combined Cars & Trucks...... 305 293 280 266 250
----------------------------------------------------------------------------------------------------------------

Several auto manufacturers stated that the increasingly stringent
requirements for fuel economy and GHG emissions in the early years of
the program should follow a more linear phase-in. The agencies'
consideration of comments and of their updated technical analyses did
not lead to changes to the phase-in of the standards discussed above.
This issue is discussed in more detail in Sections II.D, and in
Sections III and IV.
NHTSA's and EPA's technology assessment indicates there is a wide
range of technologies available for manufacturers to consider in
upgrading vehicles to reduce GHG emissions and improve fuel economy.
Commenters were in general agreement with this assessment.\28\ As
noted, these include improvements to the engines such as use of
gasoline direct injection and downsized engines that use turbochargers
to provide performance similar to that of larger engines, the use of
advanced transmissions, increased use of start-stop technology,
improvements in tire rolling resistance, reductions in vehicle weight,
increased use of hybrid and other advanced technologies, and the
initial commercialization of electric vehicles and plug-in hybrids. EPA
is also projecting improvements in vehicle air conditioners including
more efficient as well as low leak systems. All of these technologies
are already available today, and EPA's and NHTSA's assessments are that
manufacturers will be able to meet the standards through more
widespread use of these technologies across the fleet.
---------------------------------------------------------------------------

\28\ The close relationship between emissions of
CO2--the most prevalent greenhouse gas emitted by motor
vehicles--and fuel consumption, means that the technologies to
control CO2 emissions and to improve fuel economy overlap
to a great degree.
---------------------------------------------------------------------------

With respect to the practicability of the standards in terms of
lead time, during MYs 2012-2016 manufacturers are expected to go
through the normal automotive business cycle of redesigning and
upgrading their light-duty vehicle products, and in some cases
introducing entirely new vehicles not on the market today. This rule
allows manufacturers the time needed to incorporate technology to
achieve GHG reductions and improve fuel economy during the vehicle
redesign process. This is an important aspect of the rule, as it avoids
the much higher costs that would occur if manufacturers needed to add
or change technology at times other than their scheduled redesigns.
This time period also provides manufacturers the opportunity to plan
for compliance using a multi-year time frame, again consistent with
normal business practice. Over these five model years, there will be an
opportunity for manufacturers to evaluate almost every one of their
vehicle model platforms and add technology in a cost effective way to
control GHG emissions and improve fuel economy. This includes redesign
of the air conditioner systems in ways that will further reduce GHG
emissions. Various commenters stated that the proposed phase-in of the
standards should be introduced more aggressively, less aggressively, or
in a more linear manner. However, our consideration of these comments
about the phase-in, as well as our revised analyses, leads us to
conclude that the general rate of introduction of the standards as
proposed remains appropriate. This conclusion is also not affected by
the slight difference from the proposal in the final footprint-based
curves. These issues are addressed further in Sections III and IV.
Both agencies considered other standards as part of the rulemaking
analyses, both more and less stringent than those proposed. EPA's and
NHTSA's analyses of alternative standards are contained in Sections III
and IV of this preamble, respectively, as well as the agencies'
respective RIAs.
The CAFE and GHG standards described above are based on determining
emissions and fuel economy using the city and highway test procedures
that are currently used in the CAFE program. Some environmental and
other organizations commented that the test procedures should be
improved to reflect more real-world driving conditions; auto
manufacturers in general do not support such changes to the test
procedures at this time. Both agencies recognize that these test
procedures are not fully representative of real-world driving
conditions. For example, EPA has adopted more representative test
procedures that are used in determining compliance with emissions
standards for pollutants other than GHGs. These test procedures are
also used in EPA's fuel economy labeling program. However, as discussed
in Section III, the current information on effectiveness of the
individual emissions control technologies is based on performance over
the CAFE test procedures. For that reason, EPA is using the current
CAFE test procedures for the CO2 standards and is not
changing those test procedures in this rulemaking. NHTSA, as discussed
above, is limited by statute in what test procedures can be used for
purposes of passenger car testing, although there is no such statutory
limitation with respect to test procedures for trucks. However, the
same reasons for not changing the truck test procedures apply for CAFE
as well.
Both EPA and NHTSA are interested in developing programs that
employ test procedures that are more representative of real-world
driving conditions, to the extent authorized under their respective
statutes. This is an important issue, and the agencies intend to
continue to evaluate it in the context of a future rulemaking to
address standards for model year 2017 and thereafter. This could
include consideration of a range of test procedure changes to better
represent real-world driving conditions in terms of speed,
acceleration, deceleration, ambient temperatures, use of air
conditioners, and the like. With respect to air conditioner operation,
EPA discusses the public comments on these issues and the final
procedures for determining emissions credits for controls on air
conditioners in Section III.

[[Page 25333]]

Finally, based on the information EPA developed in its recent
rulemaking that updated its fuel economy labeling program to better
reflect average real-world fuel economy, the calculation of fuel
savings and CO2 emissions reductions that will be achieved
by the CAFE and GHG standards includes adjustments to account for the
difference between the fuel economy level measured in the CAFE test
procedure and the fuel economy actually achieved on average under real-
world driving conditions. These adjustments are industry averages for
the vehicles' performance as a whole, however, and are not a substitute
for the information on effectiveness of individual control technologies
that will be explored for purposes of a future GHG and CAFE rulemaking.
3. Form of the Standards
NHTSA and EPA proposed attribute-based standards for passenger cars
and light trucks. NHTSA adopted an attribute approach based on vehicle
footprint in its Reformed CAFE program for light trucks for model years
2008-2011,\29\ and recently extended this approach to passenger cars in
the CAFE rule for MY 2011 as required by EISA.\30\ The agencies also
proposed using vehicle footprint as the attribute for the GHG and CAFE
standards. Footprint is defined as a vehicle's wheelbase multiplied by
its track width--in other words, the area enclosed by the points at
which the wheels meet the ground. Most commenters that expressed a view
on this topic supported basing the standards on an attribute, and
almost all of these supported the proposed choice of vehicle footprint
as an appropriate attribute. The agencies continue to believe that the
standards are best expressed in terms of an attribute, and that the
footprint attribute is the most appropriate attribute on which to base
the standards. These issues are further discussed later in this notice
and in Chapter 2 of the Joint TSD.
---------------------------------------------------------------------------

\29\ 71 FR 17566 (Apr. 6, 2006).
\30\ 74 FR 14196 (Mar. 30, 2009).
---------------------------------------------------------------------------

Under the footprint-based standards, each manufacturer will have a
GHG and CAFE target unique to its fleet, depending on the footprints of
the vehicle models produced by that manufacturer. A manufacturer will
have separate footprint-based standards for cars and for trucks.
Generally, larger vehicles (i.e., vehicles with larger footprints) will
be subject to less stringent standards (i.e., higher CO2
grams/mile standards and lower CAFE standards) than smaller vehicles.
This is because, generally speaking, smaller vehicles are more capable
of achieving lower levels of CO2 and higher levels of fuel
economy than larger vehicles. While a manufacturer's fleet average
standard could be estimated throughout the model year based on
projected production volume of its vehicle fleet, the standard to which
the manufacturer must comply will be based on its final model year
production figures. A manufacturer's calculation of fleet average
emissions at the end of the model year will thus be based on the
production-weighted average emissions of each model in its fleet.
The final footprint-based standards are very similar in shape to
those proposed. NHTSA and EPA include more discussion of the
development of the final curves in Section II below, with a full
discussion in the Joint TSD. In addition, a full discussion of the
equations and coefficients that define the curves is included in
Section III for the CO2 curves and Section IV for the mpg
curves. The following figures illustrate the standards. First, Figure
I.B.3-1 shows the fuel economy (mpg) car standard curve.
Under an attribute-based standard, every vehicle model has a
performance target (fuel economy for the CAFE standards, and
CO2 g/mile for the GHG emissions standards), the level of
which depends on the vehicle's attribute (for this rule, footprint).
The manufacturers' fleet average performance is determined by the
production-weighted \31\ average (for CAFE, harmonic average) of those
targets. NHTSA and EPA are setting CAFE and CO2 emissions
standards defined by constrained linear functions and, equivalently,
piecewise linear functions.\32\ As a possible option for future
rulemakings, the constrained linear form was introduced by NHTSA in the
2007 NPRM proposing CAFE standards for MY 2011-2015.
---------------------------------------------------------------------------

\31\ Based on vehicles produced for sale in the United States.
\32\ The equations are equivalent but are specified differently
due to differences in the agencies' respective models.
---------------------------------------------------------------------------

NHTSA is establishing the attribute curves below for assigning a
fuel economy level to an individual vehicle's footprint value, for
model years 2012 through 2016. These mpg values will be production
weighted to determine each manufacturer's fleet average standard for
cars and trucks. Although the general model of the equation is the same
for each vehicle category and each year, the parameters of the equation
differ for cars and trucks. Each parameter also changes on an annual
basis, resulting in the yearly increases in stringency. Figure I.B.3-1
below illustrates the passenger car CAFE standard curves for model
years 2012 through 2016 while Figure I.B.3-2 below illustrates the
light truck standard curves for model years 2012-2016. The MY 2011
final standards for cars and trucks, which are specified by a
constrained logistic function rather than a constrained linear
function, are shown for comparison.
BILLING CODE 6560-50-P

[[Page 25334]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.000

[[Page 25335]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.001

[[Page 25336]]

EPA is establishing the attribute curves below for assigning a
CO2 level to an individual vehicle's footprint value, for
model years 2012 through 2016. These CO2 values will be
production weighted to determine each manufacturer's fleet average
standard for cars and trucks. As with the CAFE curves above, the
general form of the equation is the same for each vehicle category and
each year, but the parameters of the equation differ for cars and
trucks. Again, each parameter also changes on an annual basis,
resulting in the yearly increases in stringency. Figure I.B.3-3 below
illustrates the CO2 car standard curves for model years 2012
through 2016 while Figure I.B.3-4 shows the CO2 truck
standard curves for model years 2012-2016.
[GRAPHIC] [TIFF OMITTED] TR07MY10.002

[[Page 25337]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.003

BILLING CODE 6560-50-C

[[Page 25338]]

NHTSA and EPA received a number of comments about the shape of the
car and truck curves. We address these comments further in Section II.C
below as well as in Sections III and IV.
As proposed, NHTSA and EPA will use the same vehicle category
definitions for determining which vehicles are subject to the car curve
standards versus the truck curve standards. In other words, a vehicle
classified as a car under the NHTSA CAFE program will also be
classified as a car under the EPA GHG program, and likewise for trucks.
Auto industry commenters generally agreed with this approach and
believe it is an important aspect of harmonization across the two
agencies' programs. Some other commenters expressed concern about
potential consequences, especially in how cars and trucks are
distinguished. However, EPA and NHTSA are employing the same car and
truck definitions for the MY 2012-2016 CAFE and GHG standards as those
used in the CAFE program for the 2011 model year standards.\33\ This
issue is further discussed for the EPA standards in Section III, and
for the NHTSA standards in Section IV. This approach of using CAFE
definitions allows EPA's CO2 standards and the CAFE
standards to be harmonized across all vehicles for this program.
However, EPA is not changing the car/truck definition for the purposes
of any other previous rules.
---------------------------------------------------------------------------

\33\ 49 CFR 523.
---------------------------------------------------------------------------

Generally speaking, a smaller footprint vehicle will have higher
fuel economy and lower CO2 emissions relative to a larger
footprint vehicle when both have the same degree of fuel efficiency
improvement technology. In this final rule, the standards apply to a
manufacturers overall fleet, not an individual vehicle, thus a
manufacturers fleet which is dominated by small footprint vehicles will
have a higher fuel economy requirement (lower CO2
requirement) than a manufacturer whose fleet is dominated by large
footprint vehicles. A footprint-based CO2 or CAFE standard
can be relatively neutral with respect to vehicle size and consumer
choice. All vehicles, whether smaller or larger, must make improvements
to reduce CO2 emissions or improve fuel economy, and
therefore all vehicles will be relatively more expensive. With the
footprint-based standard approach, EPA and NHTSA believe there should
be no significant effect on the relative distribution of different
vehicle sizes in the fleet, which means that consumers will still be
able to purchase the size of vehicle that meets their needs. While
targets are manufacturer specific, rather than vehicle specific, Table
I.B.3-1 illustrates the fact that different vehicle sizes will have
varying CO2 emissions and fuel economy targets under the
final standards.

Table I.B.3--1 Model Year 2016 CO2 and Fuel Economy Targets for Various MY 2008 Vehicle Types
----------------------------------------------------------------------------------------------------------------
Example model
Vehicle type Example models footprint (sq. CO2 emissions Fuel economy
ft.) target (g/mi) target (mpg)
----------------------------------------------------------------------------------------------------------------
Example Passenger Cars
----------------------------------------------------------------------------------------------------------------
Compact car........................ Honda Fit............ 40 206 41.1
Midsize car........................ Ford Fusion.......... 46 230 37.1
Fullsize car....................... Chrysler 300......... 53 263 32.6
----------------------------------------------------------------------------------------------------------------
Example Light-duty Trucks
----------------------------------------------------------------------------------------------------------------
Small SUV.......................... 4WD Ford Escape...... 44 259 32.9
Midsize crossover.................. Nissan Murano........ 49 279 30.6
Minivan............................ Toyota Sienna........ 55 303 28.2
Large pickup truck................. Chevy Silverado...... 67 348 24.7
----------------------------------------------------------------------------------------------------------------

4. Program Flexibilities
EPA's and NHTSA's programs as established in this rule provide
compliance flexibility to manufacturers, especially in the early years
of the National Program. This flexibility is expected to provide
sufficient lead time for manufacturers to make necessary technological
improvements and reduce the overall cost of the program, without
compromising overall environmental and fuel economy objectives. The
broad goal of harmonizing the two agencies' standards includes
preserving manufacturers' flexibilities in meeting the standards, to
the extent appropriate and required by law. The following section
provides an overview of this final rule's flexibility provisions. Many
auto manufacturers commented in support of these provisions as critical
to meeting the standards in the lead time provided. Environmental
groups, some States, and others raised concerns about the possibility
for windfall credits and loss of program benefits. The provisions in
the final rule are in most cases the same as those proposed. However
consideration of the issues raised by commenters has led to
modifications in certain provisions. These comments and the agencies'
response are discussed in Sections III and IV below and in the Response
to Comments document.
a. CO2/CAFE Credits Generated Based on Fleet Average
Performance
Under this NHTSA and EPA final rule, the fleet average standards
that apply to a manufacturer's car and truck fleets are based on the
applicable footprint-based curves. At the end of each model year, when
production of the model year is complete, a production-weighted fleet
average will be calculated for each averaging set (cars and trucks).
Under this approach, a manufacturer's car and/or truck fleet that
achieves a fleet average CO2/CAFE level better than the
standard can generate credits. Conversely, if the fleet average
CO2/CAFE level does not meet the standard, the fleet would
incur debits (also referred to as a shortfall).
Under the final program, a manufacturer whose fleet generates
credits in a given model year would have several options for using
those credits, including credit carry-back, credit carry-forward,
credit transfers, and credit trading. These provisions exist in the MY
2011 CAFE program under EPCA and EISA, and similar provisions are part
of EPA's Tier 2 program for light-duty vehicle criteria pollutant
emissions, as well as many

[[Page 25339]]

other mobile source standards issued by EPA under the CAA. The
manufacturer will be able to carry back credits to offset a deficit
that had accrued in a prior model year and was subsequently carried
over to the current model year. EPCA also provides for this. EPCA
restricts the carry-back of CAFE credits to three years, and as
proposed EPA is establishing the same limitation, in keeping with the
goal of harmonizing both sets of standards.
After satisfying any need to offset pre-existing deficits,
remaining credits can be saved (banked) for use in future years. Under
the CAFE program, EISA allows manufacturers to apply credits earned in
a model year to compliance in any of the five subsequent model
years.\34\ As proposed, under the GHG program, EPA is also allowing
manufacturers to use these banked credits in the five years after the
year in which they were generated (i.e., five years carry-forward).
---------------------------------------------------------------------------

\34\ 49 U.S.C. 32903(a)(2).
---------------------------------------------------------------------------

EISA required NHTSA to establish by regulation a CAFE credits
transferring program, which NHTSA established in a March 2009 final
rule codified at 49 CFR Part 536, to allow a manufacturer to transfer
credits between its vehicle fleets to achieve compliance with the
standards. For example, credits earned by over-compliance with a
manufacturer's car fleet average standard could be used to offset
debits incurred due to that manufacturer's not meeting the truck fleet
average standard in a given year. EPA's Tier 2 program also provides
for this type of credit transfer. As proposed for purposes of this
rule, EPA allows unlimited credit transfers across a manufacturer's
car-truck fleet to meet the GHG standard. This is based on the
expectation that this flexibility will facilitate manufacturers'
ability to comply with the GHG standards in the lead time provided, and
will allow the required GHG emissions reductions to be achieved in the
most cost effective way. Under the CAA, unlike under EISA, there is no
statutory limitation on car-truck credit transfers. Therefore, EPA is
not constraining car-truck credit transfers, as doing so would reduce
the flexibility for lead time, and would increase costs with no
corresponding environmental benefit. For the CAFE program, however,
EISA limits the amount of credits that may be transferred, which has
the effects of limiting the extent to which a manufacturer can rely
upon credits in lieu of making fuel economy improvements to a
particular portion of its vehicle fleet, but also of potentially
increasing the costs of improving the manufacturer's overall fleet.
EISA also prohibits the use of transferred credits to meet the
statutory minimum level for the domestic car fleet standard.\35\ These
and other statutory limits will continue to apply to the determination
of compliance with the CAFE standards.
---------------------------------------------------------------------------

\35\ 49 U.S.C. 32903(g)(4).
---------------------------------------------------------------------------

EISA also allowed NHTSA to establish by regulation a CAFE credit
trading program, which NHTSA established in the March 2009 final rule
at 40 CFR part 536, to allow credits to be traded (sold) to other
vehicle manufacturers. As proposed, EPA allows credit trading in the
GHG program. These sorts of exchanges are typically allowed under EPA's
current mobile source emission credit programs, although manufacturers
have seldom made such exchanges. Under the NHTSA CAFE program, EPCA
also allows these types of credit trades, although, as with transferred
credits, traded credits may not be used to meet the minimum domestic
car standards specified by statute.\36\ Comments discussing these
provisions supported the proposed approach. These final provisions are
the same as proposed.
---------------------------------------------------------------------------

\36\ 49 U.S.C. 32903(f)(2).
---------------------------------------------------------------------------

As further discussed in Section IV of this preamble, NHTSA sought
to find a way to provide credits for improving the efficiency of light
truck air conditioners (A/Cs) and solicited public comments to that
end. The agency did so because the power necessary to operate an A/C
compressor places a significant additional load on the engine, thus
reducing fuel economy and increasing CO2 tailpipe emissions.
See Section III.C.1 below. The agency would have made a similar effort
regarding cars, but a 1975 statutory provision made it unfruitful even
to explore the possibility of administratively proving such credits for
cars. The agency did not identify a workable way of providing such
credits for light trucks in the context of this rulemaking.
b. Air Conditioning Credits Under the EPA Final Rule
Air conditioning (A/C) systems contribute to GHG emissions in two
ways. Hydrofluorocarbon (HFC) refrigerants, which are powerful GHGs,
can leak from the A/C system (direct A/C emissions). As just noted,
operation of the A/C system also places an additional load on the
engine, which results in additional CO2 tailpipe emissions
(indirect A/C related emissions). EPA is allowing manufacturers to
generate credits by reducing either or both types of GHG emissions
related to A/C systems. Specifically, EPA is establishing a method to
calculate CO2 equivalent reductions for the vehicle's full
useful life on a grams/mile basis that can be used as credits in
meeting the fleet average CO2 standards. EPA's analysis
indicates that this approach provides manufacturers with a highly cost-
effective way to achieve a portion of GHG emissions reductions under
the EPA program. EPA is estimating that manufacturers will on average
generate 11 g/mi GHG credit toward meeting the 250 g/mi by 2016 (though
some companies may generate more). EPA will also allow manufacturers to
earn early A/C credits starting in MY 2009 through 2011, as discussed
further in a later section. There were many comments on the proposed A/
C provisions. Nearly every one of these was supportive of EPA including
A/C control as part of this rule, though there was some disagreement on
some of the details of the program. The HFC crediting scheme was widely
supported. The comments mainly were concentrated on indirect A/C
related credits. The auto manufacturers and suppliers had some
technical comments on A/C technologies, and there were many concerns
with the proposed idle test. EPA has made some minor adjustments in
both of these areas that we believe are responsive to these concerns.
EPA addresses A/C issues in greater detail in Section III of this
preamble and in Chapter 2 of EPA's RIA.
c. Flexible-Fuel and Alternative Fuel Vehicle Credits
EPCA authorizes a compliance flexibility incentive under the CAFE
program for production of dual-fueled or flexible-fuel vehicles (FFV)
and dedicated alternative fuel vehicles. FFVs are vehicles that can run
both on an alternative fuel and conventional fuel. Most FFVs are E85
capable vehicles, which can run on either gasoline or a mixture of up
to 85 percent ethanol and 15 percent gasoline (E85). Dedicated
alternative fuel vehicles are vehicles that run exclusively on an
alternative fuel. EPCA was amended by EISA to extend the period of
availability of the FFV incentive, but to begin phasing it out by
annually reducing the amount of FFV incentive that can be used toward
compliance with the CAFE standards.\37\ Although NHTSA

[[Page 25340]]

expressed concern about the non-use of alternative fuel by FFVs in a
2002 report to Congress (Effects of the Alternative Motor Fuels Act
CAFE Incentives Policy), EISA does not premise the availability of the
FFV credits on actual use of alternative fuel by an FFV vehicle. Under
NHTSA's CAFE program, pursuant to EISA, no FFV credits will be
available for CAFE compliance after MY 2019.\38\ For dedicated
alternative fuel vehicles, there are no limits or phase-out of the
credits. As required by the statute, NHTSA will continue to allow the
use of FFV credits for purposes of compliance with the CAFE standards
until the end of the EISA phase-out period.
---------------------------------------------------------------------------

\37\ EPCA provides a statutory incentive for production of FFVs
by specifying that their fuel economy is determined using a special
calculation procedure that results in those vehicles being assigned
a higher fuel economy level than would otherwise occur. This is
typically referred to as an FFV credit.
\38\ Id.
---------------------------------------------------------------------------

For the GHG program, as proposed, EPA will allow FFV credits in
line with EISA limits, but only during the period from MYs 2012 to
2015. After MY 2015, EPA will only allow FFV credits based on a
manufacturer's demonstration that the alternative fuel is actually
being used in the vehicles and based on the vehicle's actual
performance. EPA discusses this in more detail in Section III.C of the
preamble, including a summary of key comments. These provisions are
being finalized as proposed, with further discussion in Section III.C
of how manufacturers can demonstrate that the alternative fuel is being
used.
d. Temporary Lead-Time Allowance Alternative Standards Under the EPA
Final Rule
Manufacturers with limited product lines may be especially
challenged in the early years of the National Program, and need
additional lead time. Manufacturers with narrow product offerings may
not be able to take full advantage of averaging or other program
flexibilities due to the limited scope of the types of vehicles they
sell. For example, some smaller volume manufacturer fleets consist
entirely of vehicles with very high baseline CO2 emissions.
Their vehicles are above the CO2 emissions target for that
vehicle footprint, but do not have other types of vehicles in their
production mix with which to average. Often, these manufacturers pay
fines under the CAFE program rather than meet the applicable CAFE
standard. EPA believes that these technological circumstances call for
more lead time in the form of a more gradual phase-in of standards.
EPA is finalizing a temporary lead-time allowance for manufacturers
that sell vehicles in the U.S. in MY 2009 and for which U.S. vehicle
sales in that model year are below 400,000 vehicles. This allowance
will be available only during the MY 2012-2015 phase-in years of the
program. A manufacturer that satisfies the threshold criteria will be
able to treat a limited number of vehicles as a separate averaging
fleet, which will be subject to a less stringent GHG standard.\39\
Specifically, a standard of 25 percent above the vehicle's otherwise
applicable foot-print target level will apply to up to 100,000 vehicles
total, spread over the four year period of MY 2012 through 2015. Thus,
the number of vehicles to which the flexibility could apply is limited.
EPA also is setting appropriate restrictions on credit use for these
vehicles, as discussed further in Section III. By MY 2016, these
allowance vehicles must be averaged into the manufacturer's full fleet
(i.e., they will no longer be eligible for a different standard). EPA
discusses this in more detail in Section III.B of the preamble.
---------------------------------------------------------------------------

\39\ EPCA does not permit such an allowance. Consequently,
manufacturers who may be able to take advantage of a lead-time
allowance under the GHG standards would be required to comply with
the applicable CAFE standard or be subject to penalties for non-
compliance.
---------------------------------------------------------------------------

EPA received comments from several smaller manufacturers that the
TLAAS program was insufficient to allow manufacturers with very limited
product lines to comply. These manufacturers commented that they need
additional lead time to meet the standards, because their
CO2 baselines are significantly higher and their vehicle
product lines are even more limited, reducing their ability to average
across their fleets compared even to other TLAAS manufacturers. EPA
fully summarizes the public comments on the TLAAS program, including
comments not supporting the program, in Section III.B. In summary, in
response to the lead time issues raised by manufacturers, EPA is
modifying the TLAAS program that applies to manufacturers with between
5,000 and 50,000 U.S. vehicle sales in MY 2009. EPA believes these
provisions are necessary given that, compared with other TLAAS
manufacturers, these manufacturers have even more limited product
offerings across which to average and higher baseline CO2
emissions, and thus need additional lead-time to meet the standards.
These manufacturers would have an increased allotment of vehicles, a
total of 250,000, compared to 100,000 vehicles (for other TLAAS-
eligible manufacturers). In addition, the TLAAS program for these
manufacturers would be extended by one year, through MY 2016 for these
vehicles, for a total of five years of eligibility. The other
provisions of the TLAAS program would continue to apply, such as the
restrictions on credit trading and the level of the standard.
Additional restrictions would also apply to these vehicles, as
discussed in Section III. In addition, for the smallest volume
manufacturers, those with below 5,000 U.S. vehicle sales, EPA is not
setting standards at this time but is instead deferring standards until
a future rulemaking. This is essentially the same approach we are using
for small businesses, which are exempted from this rule. The unique
issues involved with these manufacturers will be addressed in that
future rulemaking. Further discussion of the public comment on these
issues and details on these changes from the proposed program are
included in Section III.
e. Additional Credit Opportunities Under the Clean Air Act (CAA)
EPA is establishing additional opportunities for early credits in
MYs 2009-2011 through over-compliance with a baseline standard. The
baseline standard is set to be equivalent, on a national level, to the
California standards. Credits can be generated by over-compliance with
this baseline in one of two ways--over-compliance by the fleet of
vehicles sold in California and the CAA section 177 States (i.e., those
States adopting the California program), or over-compliance with the
fleet of vehicles sold in the 50 States. EPA is also providing for
early credits based on over-compliance with CAFE, but only for vehicles
sold in States outside of California and the CAA section 177 states.
Under the early credit provisions, no early FFV credits would be
allowed, except those achieved by over-compliance with the California
program based on California's provisions that manufacturers demonstrate
actual use of the alternative fuel. EPA's early credits provisions are
designed to ensure that there would be no double counting of early
credits. NHTSA notes, however, that credits for overcompliance with
CAFE standards during MYs 2009-2011 will still be available for
manufacturers to use toward compliance in future model years, just as
before.
EPA received comments from some environmental organizations and
States expressing concern that these early credits were inappropriate
windfall credits because they provided credits for actions that were
not surplus, that is above what would otherwise be required for
compliance with either State or Federal motor vehicle standards. This
focused on the credits

[[Page 25341]]

for over-compliance with the California standards generated during
model years 2009 and perhaps 2010, where according to commenters the
CAFE requirements were in effect more stringent than the California
standards. EPA believes that early credits provide a valuable incentive
for manufacturers that have implemented fuel efficient technologies in
excess of their CAFE compliance obligations prior to MY 2012. With
appropriate restrictions, these credits, reflecting over-compliance
over a three model year time frame (MY 2009-2011) and not just over one
or two model years, will be surplus reductions and not otherwise
required by law. Therefore, EPA is finalizing these provisions largely
as proposed, but in response to comments, with an additional
restriction on the trading of MY 2009 credits. The overall structure of
this early credit program addresses concerns about the potential for
windfall credits in the first one or two model years. This issue is
fully discussed in Section III.C.
EPA is providing an additional temporary incentive to encourage the
commercialization of advanced GHG/fuel economy control technologies--
including electric vehicles (EVs), plug-in hybrid electric vehicles
(PHEVs), and fuel cell vehicles (FCVs)--for model years 2012-2016.
EPA's proposal included an emissions compliance value of zero grams/
mile for EVs and FCVs, and the electric portion of PHEVs, and a
multiplier in the range of 1.2 to 2.0, so that each advanced technology
vehicle would count as greater than one vehicle in a manufacturer's
fleetwide compliance calculation. EPA received many comments on the
proposed incentives. Many State and environmental organization
commenters believed that the combination of these incentives could
undermine the GHG benefits of the rule, and believed the emissions
compliance values should take into account the net upstream GHG
emissions associated with electrified vehicles compared to vehicles
powered by petroleum based fuel. Auto manufacturers generally supported
the incentives, some believing the incentives to be a critical part of
the National Program. Most auto makers supported both the zero grams/
mile emissions compliance value and the higher multipliers.
Upon considering the public comments on this issue, EPA is
finalizing an advanced technology vehicle incentive program that
includes a zero gram/mile emissions compliance value for EVs and FCVs,
and the electric portion of PHEVs, for up to the first 200,000 EV/PHEV/
FCV vehicles produced by a given manufacturer during MY 2012-2016 (for
a manufacturer that produces less than 25,000 EVs, PHEVs, and FCVs in
MY 2012), or for up to the first 300,000 EV/PHEV/FCV vehicles produced
during MY 2012-2016 (for a manufacturer that produces 25,000 or more
EVs, PHEVs, and FCVs in MY 2012). For any production greater than this
amount, the compliance value for the vehicle will be greater than zero
gram/mile, set at a level that reflects the vehicle's net increase in
upstream GHG emissions in comparison to the gasoline vehicle it
replaces. In addition, EPA is not finalizing a multiplier. EPA will
also allow this early advanced technology incentive program beginning
in MYs 2009-2011. The purpose of these provisions is to provide a
temporary incentive to promote technologies which have the potential to
produce very large GHG reductions in the future. The tailpipe GHG
emissions from EVs, FCVs, and PHEVs operated on grid electricity are
zero, and traditionally the emissions of the vehicle itself are all
that EPA takes into account for purposes of compliance with standards
set under section 202(a). This has not raised any issues for criteria
pollutants, as upstream emissions associated with production and
distribution of the fuel are addressed by comprehensive regulatory
programs focused on the upstream sources of those emissions. At this
time, however, there is no such comprehensive program addressing
upstream emissions of GHGs, and the upstream GHG emissions associated
with production and distribution of electricity are higher than the
corresponding upstream GHG emissions of gasoline or other petroleum
based fuels. In the future, vehicle fleet electrification combined with
advances in low-carbon technology in the electricity sector have the
potential to transform the transportation sector's contribution to the
country's GHG emissions. EPA will reassess the issue of how to address
EVs, PHEVs, and FCVs in rulemakings for model years 2017 and beyond,
based on the status of advanced vehicle technology commercialization,
the status of upstream GHG control programs, and other relevant
factors. Further discussion of the temporary advanced technology
vehicle incentives, including more detail on the public comments and
EPA's response, is found in Section III.C.
EPA is also providing an option for manufacturers to generate
credits for employing new and innovative technologies that achieve GHG
reductions that are not reflected on current test procedures, as
proposed. Examples of such ``off-cycle'' technologies might include
solar panels on hybrids, adaptive cruise control, and active
aerodynamics, among other technologies. These three credit provisions
are discussed in more detail in Section III.
5. Coordinated Compliance
Previous NHTSA and EPA regulations and statutory provisions
establish ample examples on which to develop an effective compliance
program that achieves the energy and environmental benefits from CAFE
and motor vehicle GHG standards. NHTSA and EPA have developed a program
that recognizes, and replicates as closely as possible, the compliance
protocols associated with the existing CAA Tier 2 vehicle emission
standards, and with CAFE standards. The certification, testing,
reporting, and associated compliance activities closely track current
practices and are thus familiar to manufacturers. EPA already oversees
testing, collects and processes test data, and performs calculations to
determine compliance with both CAFE and CAA standards. Under this
coordinated approach, the compliance mechanisms for both programs are
consistent and non-duplicative. EPA will also apply the CAA authorities
applicable to its separate in-use requirements in this program.
The compliance approach allows manufacturers to satisfy the new
program requirements in the same general way they comply with existing
applicable CAA and CAFE requirements. Manufacturers would demonstrate
compliance on a fleet-average basis at the end of each model year,
allowing model-level testing to continue throughout the year as is the
current practice for CAFE determinations. The compliance program design
establishes a single set of manufacturer reporting requirements and
relies on a single set of underlying data. This approach still allows
each agency to assess compliance with its respective program under its
respective statutory authority.
NHTSA and EPA do not anticipate any significant noncompliance under
the National Program. However, failure to meet the fleet average
standards (after credit opportunities are exhausted) would ultimately
result in the potential for penalties under both EPCA and the CAA. The
CAA allows EPA considerable discretion in assessment of penalties.
Penalties under the CAA are typically determined on a vehicle-specific
basis by determining the

[[Page 25342]]

number of a manufacturer's highest emitting vehicles that caused the
fleet average standard violation. This is the same mechanism used for
EPA's National Low Emission Vehicle and Tier 2 corporate average
standards, and to date there have been no instances of noncompliance.
CAFE penalties are specified by EPCA and would be assessed for the
entire noncomplying fleet at a rate of $5.50 times the number of
vehicles in the fleet, times the number of tenths of mpg by which the
fleet average falls below the standard. In the event of a compliance
action arising out of the same facts and circumstances, EPA could
consider CAFE penalties when determining appropriate remedies for the
EPA case.
Several stakeholders commented on the proposed coordinated
compliance approach. The comments indicated broad support for the
overall approach EPA proposed. In particular, both regulated industry
and the public interest community appreciated the attempt to streamline
compliance by adopting current practice where possible and by
coordinating EPA and NHTSA compliance requirements. Thus the final
compliance program design is largely unchanged from the proposal. Some
commenters requested additional detail or clarification in certain
areas and others suggested some relatively narrow technical changes,
and EPA has responded to these suggestions. EPA and NHTSA summarize
these comments and the agencies' responses in Sections III and IV,
respectively, below. The Response to Comments document associated with
this document includes all of the comments and responses received
during the comment period.

C. Summary of Costs and Benefits of the National Program

This section summarizes the projected costs and benefits of the
CAFE and GHG emissions standards. These projections helped inform the
agencies' choices among the alternatives considered and provide further
confirmation that the final standards are an appropriate choice within
the spectrum of choices allowable under their respective statutory
criteria. The costs and benefits projected by NHTSA to result from
these CAFE standards are presented first, followed by those from EPA's
analysis of the GHG emissions standards.
For several reasons, the estimates for costs and benefits presented
by NHTSA and EPA, while consistent, are not directly comparable, and
thus should not be expected to be identical. Most important, NHTSA and
EPA's standards would require slightly different fuel efficiency
improvements. EPA's GHG standard is more stringent in part due to its
assumptions about manufacturers' use of air conditioning credits, which
result from reductions in air conditioning-related emissions of HFCs
and CO2. NHTSA was unable to make assumptions about
manufacturers' improving the efficiency of air conditioners due to
statutory limitations. In addition, the CAFE and GHG standards offer
different program flexibilities, and the agencies' analyses differ in
their accounting for these flexibilities (for example, FFVs), primarily
because NHTSA is statutorily prohibited from considering some
flexibilities when establishing CAFE standards, while EPA is not. These
differences contribute to differences in the agencies' respective
estimates of costs and benefits resulting from the new standards.
NHTSA performed two analyses: a primary analysis that shows the
estimates of costs, fuel savings, and related benefits that the agency
considered for purposes of establishing new CAFE standards, and a
supplemental analysis that reflects the agency's best estimate of the
potential real-world effects of the CAFE standards, including
manufacturers' potential use of FFV credits in accordance with the
provisions of EISA concerning their availability. Because EPCA
prohibits NHTSA from considering the ability of manufacturers to use of
FFV credits to increase their fleet average fuel economy when
establishing CAFE standards, the agency's primary analysis does not
include them. However, EPCA does not prohibit NHTSA from considering
the fact that manufacturers may pay civil penalties rather than
complying with CAFE standards, and NHTSA's primary analysis accounts
for some manufacturers' tendency to do so. In addition, NHTSA's
supplemental analysis of the effect of FFV credits on benefits and
costs from its CAFE standards, demonstrates the real-world impacts of
FFVs, and the summary estimates presented in Section IV include these
effects. Including the use of FFV credits reduces estimated per-vehicle
compliance costs of the program. However, as shown below, including FFV
credits does not significantly change the projected fuel savings and
CO2 reductions, because FFV credits reduce the fuel economy
levels that manufacturers achieve not only under the standards, but
also under the baseline MY 2011 CAFE standards.
Also, EPCA, as amended by EISA, allows manufacturers to transfer
credits between their passenger car and light truck fleets. However,
EPCA also prohibits NHTSA from considering manufacturers' ability to
increase their average fuel economy through the use of CAFE credits
when determining the stringency of the CAFE standards. Because of this
prohibition, NHTSA's primary analysis does not account for the extent
to which credit transfers might actually occur. For purposes of its
supplemental analysis, NHTSA considered accounting for the possibility
that some manufacturers might utilize the opportunity under EPCA to
transfer some CAFE credits between the passenger car and light truck
fleets, but determined that in NHTSA's year-by-year analysis,
manufacturers' credit transfers cannot be reasonably estimated at this
time.\40\
---------------------------------------------------------------------------

\40\ NHTSA's analysis estimates multi-year planning effects
within a context in which each model year is represented explicitly,
and technologies applied in one model year carry forward to future
model years. NHTSA does not currently have a reasonable basis to
estimate how a manufacturer might, for example, weigh the transfer
of credits from the passenger car to the light truck fleet in MY
2013 against the potential to carry light truck technologies forward
from MY 2013 through MY 2016.
---------------------------------------------------------------------------

EPA made explicit assumptions about manufacturers' use of FFV
credits under both the baseline and control alternatives, and its
estimates of costs and benefits from the GHG standards reflect these
assumptions. However, under the GHG standards, FFV credits would be
available through MY 2015; starting in MY 2016, EPA will only allow FFV
credits based on a manufacturer's demonstration that the alternative
fuel is actually being used in the vehicles and the actual GHG
performance for the vehicle run on that alternative fuel.
EPA's analysis also assumes that manufacturers would transfer
credits between their car and truck fleets in the MY 2011 baseline
subject to the maximum value allowed by EPCA, and that unlimited car-
truck credit transfers would occur under the GHG standards. Including
these assumptions in EPA's analysis increases the resulting estimates
of fuel savings and reductions in GHG emissions, while reducing EPA's
estimates of program compliance costs.
Finally, under the EPA GHG program, there is no ability for a
manufacturer to intentionally pay fines in lieu of meeting the
standard. Under EPCA, however, vehicle manufacturers are allowed to pay
fines as an alternative to compliance with applicable CAFE standards.
NHTSA's analysis explicitly estimates the level of voluntary fine
payment by individual manufacturers, which reduces NHTSA's estimates of

[[Page 25343]]

both the costs and benefits of its CAFE standards. In contrast, the CAA
does not allow for fine payment (civil penalties) in lieu of compliance
with emission standards, and EPA's analysis of benefits from its
standard thus assumes full compliance. This assumption results in
higher estimates of fuel savings, of reductions in GHG emissions, and
of manufacturers' compliance costs to sell fleets that comply with both
NHTSA's CAFE program and EPA's GHG program.
In summary, the projected costs and benefits presented by NHTSA and
EPA are not directly comparable, because the GHG emission levels
established by EPA include air conditioning-related improvements in
equivalent fuel efficiency and HFC reductions, because of the
assumptions incorporated in EPA's analysis regarding car-truck credit
transfers, and because of EPA's projection of complete compliance with
the GHG standards. It should also be expected that overall, EPA's
estimates of GHG reductions and fuel savings achieved by the GHG
standards will be slightly higher than those projected by NHTSA only
for the CAFE standards because of the reasons described above. For the
same reasons, EPA's estimates of manufacturers' costs for complying
with the passenger car and light trucks GHG standards are slightly
higher than NHTSA's estimates for complying with the CAFE standards.
A number of stakeholders commented on NHTSA's and EPA's analytical
assumptions in estimating costs and benefits of the program. These
comments and any changes from the proposed values are summarized in
Section II.F, and further in Sections III (for EPA) and IV (for NHTSA);
the Response to Comments document presents the detailed responses to
each of the comments.
1. Summary of Costs and Benefits of NHTSA's CAFE Standards
NHTSA has analyzed in detail the costs and benefits of the final
CAFE standards. Table I.C.1-1 presents the total costs, benefits, and
net benefits for NHTSA's final CAFE standards. The values in Table
I.C.1-1 display the total costs for all MY 2012-2016 vehicles and the
benefits and net benefits represent the impacts of the standards over
the full lifetime of the vehicles projected to be sold during model
years 2012-2016. It is important to note that there is significant
overlap in costs and benefits for NHTSA's CAFE program and EPA's GHG
program and therefore combined program costs and benefits, which
together comprise the National Program, are not a sum of the two
individual programs.

Table I.C.1-1--NHTSA's Estimated 2012-2016 Model Year Costs, Benefits,
and Net Benefits Under the CAFE Standards Before FFV Credits
[2007 dollars]
------------------------------------------------------------------------
3% Discount Rate: $billions
------------------------------------------------------------------------

Costs..................................................... 51.8
Benefits.................................................. 182.5
Net Benefits.............................................. 130.7
7% Discount Rate:
Costs..................................................... 51.8
Benefits.................................................. 146.3
Net Benefits.............................................. 94.5
------------------------------------------------------------------------

NHTSA estimates that these new CAFE standards will lead to fuel
savings totaling 61 billion gallons throughout the useful lives of
vehicles sold in MYs 2012-2016. At a 3% discount rate, the present
value of the economic benefits resulting from those fuel savings is
$143 billion. At a 7% discount rate, the present value of the economic
benefits resulting from those fuel savings is $112 billion.\41\
---------------------------------------------------------------------------

\41\ These figures do not account for the compliance
flexibilities that NHTSA is prohibited from considering when
determining the level of new CAFE standards, because manufacturers'
decisions to use those flexibilities are voluntary.
---------------------------------------------------------------------------

The agency further estimates that these new CAFE standards will
lead to corresponding reductions in CO2 emissions totaling
655 million metric tons (mmt) during the useful lives of vehicles sold
in MYs 2012-2016. The present value of the economic benefits from
avoiding those emissions is $14.5 billion, based on a global social
cost of carbon value of approximately $21 per metric ton (in 2010, and
growing thereafter).\42\ It is important to note that NHTSA's CAFE
standards and EPA's GHG standards will both be in effect, and each will
lead to increases in average fuel economy and CO2 emissions
reductions. The two agencies' standards together comprise the National
Program, and this discussion of costs and benefits of NHTSA's CAFE
standards does not change the fact that both the CAFE and GHG
standards, jointly, are the source of the benefits and costs of the
National Program.
---------------------------------------------------------------------------

\42\ NHTSA also estimated the benefits associated with three
more estimates of a one ton GHG reduction in 2010 ($5, $35, and
$65), which will likewise grow thereafter. See Section II for a more
detailed discussion of the social cost of carbon.

Table I.C.1-2--NHTSA Fuel Saved (Billion Gallons) and CO2 Emissions Avoided (mmt) Under CAFE Standards (Without FFV Credits)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Fuel (b. gal.).................................... 4.2 8.9 12.5 16.0 19.5 61.0
CO2 (mmt)......................................... 44 94 134 172 210 655
--------------------------------------------------------------------------------------------------------------------------------------------------------

Considering manufacturers' ability to earn credit toward compliance
by selling FFVs, NHTSA estimates very little change in incremental fuel
savings and avoided CO2 emissions, assuming FFV credits
would be used toward both the baseline and final standards:

Table I.C.1-3--NHTSA Fuel Saved (Billion gallons) and CO2 Emissions Avoided (Million Metric Tons, mmt) Under CAFE Standards (With FFV Credits)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Fuel (b. gal.).......................................... 4.9 8.2 11.3 15.0 19.1 58.6

[[Page 25344]]

CO2 (mmt)............................................... 53 89 123 163 208 636
--------------------------------------------------------------------------------------------------------------------------------------------------------

NHTSA estimates that these fuel economy increases would produce
other benefits both to drivers (e.g., reduced time spent refueling) and
to the U.S. (e.g., reductions in the costs of petroleum imports beyond
the direct savings from reduced oil purchases, as well as some
disbenefits (e.g., increase traffic congestion) caused by drivers'
tendency to travel more when the cost of driving declines (as it does
when fuel economy increases). NHTSA has estimated the total monetary
value to society of these benefits and disbenefits, and estimates that
the standards will produce significant net benefits to society. Using a
3% discount rate, NHTSA estimates that the present value of these
benefits would total more than $180 billion over the useful lives of
vehicles sold during MYs 2012-2016. More discussion regarding monetized
benefits can be found in Section IV of this notice and in NHTSA's
Regulatory Impact Analysis. Note that the benefit calculation in Tables
I.C.1-4 through 1-7 includes the benefits of reducing CO2
emissions,\43\ but not the benefits of reducing other GHG emissions.
---------------------------------------------------------------------------

\43\ CO2 benefits for purposes of these tables are
calculated using the $21/ton SCC values. Note that net present value
of reduced GHG emissions is calculated differently than other
benefits. The same discount rate used to discount the value of
damages from future emissions (SCC at 5, 3, and 2.5 percent) is used
to calculate net present value of SCC for internal consistency.

Table I.C.1-4--NHTSA Discounted Benefits ($billion) Under the CAFE Standards (Before FFV Credits, Using 3 Percent Discount Rate)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 6.8 15.2 21.6 28.7 35.2 107.5
Light Trucks............................................ 5.1 10.7 15.5 19.4 24.3 75.0
-----------------------------------------------------------------------------------------------
Combined............................................ 11.9 25.8 37.1 48.0 59.5 182.5
--------------------------------------------------------------------------------------------------------------------------------------------------------

Using a 7% discount rate, NHTSA estimates that the present value of
these benefits would total more than $145 billion over the same time
period.

Table I.C.1-5--NHTSA Discounted Benefits ($billion) Under the CAFE Standards (Before FFV Credits, Using 7 Percent Discount Rate)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 5.5 12.3 17.5 23.2 28.6 87.0
Light Trucks............................................ 4.0 8.4 12.2 15.3 19.2 59.2
-----------------------------------------------------------------------------------------------
Combined............................................ 9.5 20.7 29.7 38.5 47.8 146.2
--------------------------------------------------------------------------------------------------------------------------------------------------------

NHTSA estimates that FFV credits could reduce achieved benefits by
about 3.8%:

Table I.C.1-6a--NHTSA Discounted Benefits ($billion) Under the CAFE Standards (With FFV Credits, Using a 3 Percent Discount Rate)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 7.6 13.7 19.1 25.6 34.0 100.0
Light Trucks............................................ 6.4 10.4 14.6 19.8 24.4 75.6
-----------------------------------------------------------------------------------------------
Combined............................................ 14.0 24.1 33.7 45.4 58.4 175.6
--------------------------------------------------------------------------------------------------------------------------------------------------------

Table I.C.1-6b--NHTSA Discounted Benefits ($billion) Under the CAFE Standards (With FFV Credits, Using a 7 Percent Discount Rate)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 6.1 11.1 15.5 20.7 27.6 80.9
Light Trucks............................................ 5.0 8.2 11.5 15.6 19.3 59.7
-----------------------------------------------------------------------------------------------

[[Page 25345]]

Combined............................................ 11.2 19.3 27.0 36.4 46.9 140.7
--------------------------------------------------------------------------------------------------------------------------------------------------------

NHTSA attributes most of these benefits--about $143 billion (at a
3% discount rate and excluding consideration of FFV credits), as noted
above--to reductions in fuel consumption, valuing fuel (for societal
purposes) at the future pre-tax prices projected in the Energy
Information Administration's (AEO's) reference case forecast from the
Annual Energy Outlook (AEO) 2010 Early Release. NHTSA's Final
Regulatory Impact Analysis (FRIA) accompanying this rule presents a
detailed analysis of specific benefits of the rule.

Table I.C.1-7--Summary of Benefits Fuel Savings and CO2 Emissions Reduction Due to the Rule (Before FFV Credits)
----------------------------------------------------------------------------------------------------------------
Monetized value (discounted)
Amount ------------------------------------------------------
3% discount rate 7% discount rate
----------------------------------------------------------------------------------------------------------------
Fuel savings...................... 61.0 billion gallons. $143.0 billion...... $112.0 billion.
CO2 emissions reductions.......... 655 mmt.............. $14.5 billion....... $14.5 billion.
----------------------------------------------------------------------------------------------------------------

NHTSA estimates that the increases in technology application
necessary to achieve the projected improvements in fuel economy will
entail considerable monetary outlays. The agency estimates that
incremental costs for achieving its standards--that is, outlays by
vehicle manufacturers over and above those required to comply with the
MY 2011 CAFE standards--will total about $52 billion (i.e., during MYs
2012-2016).

Table I.C.1-8--NHTSA Incremental Technology Outlays ($billion) Under the CAFE Standards (Before FFV Credits)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 4.1 5.4 6.9 8.2 9.5 34.2
Light Trucks............................................ 1.8 2.5 3.7 4.3 5.4 17.6
-----------------------------------------------------------------------------------------------
Combined............................................ 5.9 7.9 10.5 12.5 14.9 51.7
--------------------------------------------------------------------------------------------------------------------------------------------------------

NHTSA estimates that use of FFV credits could significantly reduce
these outlays:

Table I.C.1-9--NHTSA Incremental Technology Outlays ($billion) under CAFE Standards (With FFV Credits)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Passenger Cars.......................................... 2.6 3.6 4.8 6.1 7.5 24.6
Light Trucks............................................ 1.1 1.5 2.5 3.4 4.4 12.9
-----------------------------------------------------------------------------------------------
Combined............................................ 3.7 5.1 7.3 9.5 11.9 37.5
--------------------------------------------------------------------------------------------------------------------------------------------------------

The agency projects that manufacturers will recover most or all of
these additional costs through higher selling prices for new cars and
light trucks. To allow manufacturers to recover these increased outlays
(and, to a much lesser extent, the civil penalties that some companies
are expected to pay for noncompliance), the agency estimates that the
standards would lead to increases in average new vehicle prices ranging
from $457 per vehicle in MY 2012 to $985 per vehicle in MY 2016:

Table I.C.1-10--NHTSA Incremental Increases in Average New Vehicle Costs ($) Under CAFE Standards (Before FFV
Credits)
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................. 505 573 690 799 907
Light Trucks.................... 322 416 621 752 961
-------------------------------------------------------------------------------

[[Page 25346]]

Combined.................... 434 513 665 782 926
----------------------------------------------------------------------------------------------------------------

NHTSA estimates that use of FFV credits could significantly reduce
these costs, especially in earlier model years:

Table I.C.1-11--NHTSA Incremental Increases in Average New Vehicle Costs ($) Under CAFE Standards (With FFV
Credits)
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................. 303 378 481 593 713
Light Trucks.................... 194 260 419 581 784
-------------------------------------------------------------------------------
Combined.................... 261 333 458 589 737
----------------------------------------------------------------------------------------------------------------

NHTSA estimates, therefore, that the total benefits of these CAFE
standards will be more than three times the magnitude of the
corresponding costs. As a consequence, its standards would produce net
benefits of $130.7 billion at a 3 percent discount rate (with FFV
credits, $138.2 billion) or $94.5 billion at a 7 percent discount rate
over the useful lives of vehicles sold during MYs 2012-2016.
2. Summary of Costs and Benefits of EPA's GHG Standards
EPA has analyzed in detail the costs and benefits of the final GHG
standards. Table I.C.2-1 shows EPA's estimated lifetime discounted
cost, benefits and net benefits for all vehicles projected to be sold
in model years 2012-2016. It is important to note that there is
significant overlap in costs and benefits for NHTSA's CAFE program and
EPA's GHG program and therefore combined program costs and benefits are
not a sum of the individual programs.

Table I.C.2-1--EPA's Estimated 2012-2016 Model Year Lifetime Discounted
Costs, Benefits, and Net Benefits Assuming the $21/Ton SCC Value a b c d
[2007 dollars]
------------------------------------------------------------------------
3% Discount rate $Billions
------------------------------------------------------------------------
Costs..................................................... 51.5
Benefits.................................................. 240
Net Benefits.............................................. 189
------------------------------------------------------------------------
7% Discount rate
------------------------------------------------------------------------
Costs..................................................... 51.5
Benefits.................................................. 192
Net Benefits.............................................. 140
------------------------------------------------------------------------
\a\ Although EPA estimated the benefits associated with four different
values of a one ton GHG reduction ($5, $21, $35, $65), for the
purposes of this overview presentation of estimated costs and benefits
EPA is showing the benefits associated with the marginal value deemed
to be central by the interagency working group on this topic: $21 per
ton of CO2e, in 2007 dollars and 2010 emissions. The $21/ton value
applies to 2010 CO2 emissions and grows over time.
\b\ As noted in Section III.H, SCC increases over time. The $21/ton
value applies to 2010 CO2 emissions and grows larger over time.
\c\ Note that net present value of reduced GHG emissions is calculated
differently than other benefits. The same discount rate used to
discount the value of damages from future emissions (SCC at 5, 3, and
2.5 percent) is used to calculate net present value of SCC for
internal consistency. Refer to Section III.H for more detail.
\d\ Monetized GHG benefits exclude the value of reductions in non-CO2
GHG emissions (HFC, CH4 and N2O) expected under this final rule.
Although EPA has not monetized the benefits of reductions in these non-
CO2 emissions, the value of these reductions should not be interpreted
as zero. Rather, the reductions in non-CO2 GHGs will contribute to
this rule's climate benefits, as explained in Section III.F.2. The SCC
TSD notes the difference between the social cost of non-CO2 emissions
and CO2 emissions, and specifies a goal to develop methods to value
non-CO2 emissions in future analyses.

Table I.C.2-2 shows EPA's estimated lifetime fuel savings and
CO2 equivalent emission reductions for all vehicles sold in
the model years 2012-2016. The values in Table I.C.2-2 are projected
lifetime totals for each model year and are not discounted. As
documented in EPA's Final RIA, the potential credit transfer between
cars and trucks may change the distribution of the fuel savings and GHG
emission impacts between cars and trucks. As discussed above with
respect to NHTSA's CAFE standards, it is important to note that NHTSA's
CAFE standards and EPA's GHG standards will both be in effect, and each
will lead to increases in average fuel economy and reductions in
CO2 emissions. The two agencies' standards together comprise
the National Program, and this discussion of costs and benefits of
EPA's GHG standards does not change the fact that both the CAFE and GHG
standards, jointly, are the source of the benefits and costs of the
National Program.

Table I.C.2-2--EPA's Estimated 2012-2016 Model Year Lifetime Fuel Saved and GHG Emissions Avoided
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
----------------------------------------------------------------------------------------------------------------
Cars.................. Fuel (billion 4.0 5.5 7.3 10.5 14.3 41.6
gallons).
Fuel (billion 0.10 0.13 0.17 0.25 0.34 0.99
barrels).
CO2 EQ (mmt).... 49.3 68.5 92.7 134 177 521

[[Page 25347]]

Light Trucks.......... Fuel (billion 3.3 5.0 6.6 9.0 12.2 36.1
gallons).
Fuel (billion 0.08 0.12 0.16 0.21 0.29 0.86
barrels).
CO2 EQ (mmt).... 39.6 61.7 81.6 111 147 441
-----------------------------------------------------------------------------------------
Combined.......... Fuel (billion 7.3 10.5 13.9 19.5 26.5 77.7
gallons).
Fuel (billion 0.17 0.25 0.33 0.46 0.63 1.85
barrels).
CO2 EQ (mmt).... 88.8 130 174 244 325 962
----------------------------------------------------------------------------------------------------------------

Table I.C.2-3 shows EPA's estimated lifetime discounted benefits
for all vehicles sold in model years 2012-2016. Although EPA estimated
the benefits associated with four different values of a one ton GHG
reduction ($5, $21, $35, $65), for the purposes of this overview
presentation of estimated benefits EPA is showing the benefits
associated with one of these marginal values, $21 per ton of
CO2, in 2007 dollars and 2010 emissions. Table I.C.2-3
presents benefits based on the $21 value. Section III.H presents the
four marginal values used to estimate monetized benefits of GHG
reductions and Section III.H presents the program benefits using each
of the four marginal values, which represent only a partial accounting
of total benefits due to omitted climate change impacts and other
factors that are not readily monetized. The values in the table are
discounted values for each model year of vehicles throughout their
projected lifetimes. The benefits include all benefits considered by
EPA such as fuel savings, GHG reductions, PM benefits, energy security
and other externalities such as reduced refueling and accidents,
congestion and noise. The lifetime discounted benefits are shown for
one of four different social cost of carbon (SCC) values considered by
EPA. The values in Table I.C.2-3 do not include costs associated with
new technology required to meet the GHG standard.

Table I.C.2-3--EPA's Estimated 2012-2016 Model Year Lifetime Discounted Benefits Assuming the $21/Ton SCC Value a b c
[Billions of 2007 dollars]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Model year
Discount rate -----------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
3%...................................................... $21.8 $32.0 $42.8 $60.8 $83.3 $240
7%...................................................... 17.4 25.7 34.2 48.6 66.4 192
--------------------------------------------------------------------------------------------------------------------------------------------------------
\a\ The benefits include all benefits considered by EPA such as the economic value of reduced fuel consumption and accompanying savings in refueling
time, climate-related economic benefits from reducing emissions of CO2 (but not other GHGs), economic benefits from reducing emissions of PM and other
air pollutants that contribute to its formation, and reductions in energy security externalities caused by U.S. petroleum consumption and imports. The
analysis also includes disbenefits stemming from additional vehicle use, such as the economic damages caused by accidents, congestion and noise.
\b\ Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same discount rate used to discount the
value of damages from future emissions (SCC at 5, 3, and 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer to
Section III.H for more detail.
\c\ Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected under this final rule. Although EPA has
not monetized the benefits of reductions in these non-CO2 emissions, the value of these reductions should not be interpreted as zero. Rather, the
reductions in non-CO2 GHGs will contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference between
the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses. Also,
as noted in Section III.H, SCC increases over time. The $21/ton value applies to 2010 emissions and grows larger over time.

Table I.C.2-4 shows EPA's estimated lifetime fuel savings, lifetime
CO2 emission reductions, and the monetized net present
values of those fuel savings and CO2 emission reductions.
The gallons of fuel and CO2 emission reductions are
projected lifetime values for all vehicles sold in the model years
2012-2016. The estimated fuel savings in billions of barrels and the
GHG reductions in million metric tons of CO2 shown in Table
I.C.2-4 are totals for the five model years throughout their projected
lifetime and are not discounted. The monetized values shown in Table
I.C.2-4 are the summed values of the discounted monetized-fuel savings
and monetized-CO2 reductions for the five model years 2012-
2016 throughout their lifetimes. The monetized values in Table I.C.2-4
reflect both a 3 percent and a 7 percent discount rate as noted.

Table I.C.2-4--EPA's Estimated 2012-2016 Model Year Lifetime Fuel Savings, CO2 Emission Reductions, and
Discounted Monetized Benefits at a 3% Discount Rate
[Monetized values in 2007 dollars]
----------------------------------------------------------------------------------------------------------------
Amount $ value (billions)
----------------------------------------------------------------------------------------------------------------
Fuel savings............................ 1.8 billion barrels........ $182, 3% discount rate.
$142, 7% discount rate.

[[Page 25348]]

CO2e emission reductions (CO2 portion 962 MMT CO2e............... $17 a b.
valued assuming $21/ton CO2 in 2010).
----------------------------------------------------------------------------------------------------------------
\a\ $17 billion for 858 MMT of reduced CO2 emissions. As noted in Section III.H, the $21/ton value applies to
2010 emissions and grows larger over time. Monetized GHG benefits exclude the value of reductions in non-CO2
GHG emissions (HFC, CH4 and N2O) expected under this final rule. Although EPA has not monetized the benefits
of reductions in these non-CO2 emissions, the value of these reductions should not be interpreted as zero.
Rather, the reductions in non-CO2 GHGs will contribute to this rule's climate benefits, as explained in
Section III.F.2. The SCC TSD notes the difference between the social cost of non-CO2 emissions and CO2
emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses.
\b\ Note that net present value of reduced CO2 emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, and 2.5 percent) is
used to calculate net present value of SCC for internal consistency. Refer to Section III.H for more detail.

Table I.C.2-5 shows EPA's estimated incremental and total
technology outlays for cars and trucks for each of the model years
2012-2016. The technology outlays shown in Table I.C.2-5 are for the
industry as a whole and do not account for fuel savings associated with
the program.

Table I.C.2-5--EPA's Estimated Incremental Technology Outlays
[Billions of 2007 dollars]
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Cars.................................................... $3.1 $5.0 $6.5 $8.0 $9.4 $31.9
Trucks.................................................. 1.8 3.0 3.9 4.8 6.2 19.7
-----------------------------------------------------------------------------------------------
Combined............................................ 4.9 8.0 10.3 12.7 15.6 51.5
--------------------------------------------------------------------------------------------------------------------------------------------------------

Table I.C.2-6 shows EPA's estimated incremental cost increase of
the average new vehicle for each model year 2012-2016. The values shown
are incremental to a baseline vehicle and are not cumulative. In other
words, the estimated increase for 2012 model year cars is $342 relative
to a 2012 model year car absent the National Program. The estimated
increase for a 2013 model year car is $507 relative to a 2013 model
year car absent the National Program (not $342 plus $507).

Table I.C.2-6--EPA's Estimated Incremental Increase in Average New Vehicle Cost
[2007 dollars per unit]
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Cars............................ $342 $507 $631 $749 $869
Trucks.......................... 314 496 652 820 1,098
-------------------------------------------------------------------------------
Combined.................... 331 503 639 774 948
----------------------------------------------------------------------------------------------------------------

D. Background and Comparison of NHTSA and EPA Statutory Authority

Section I.C of the proposal contained a detailed overview
discussion of the NHTSA and EPA statutory authorities. In addition to
the discussion in the proposal, each agency discusses comments
pertaining to its statutory authority and the agency's responses in
Sections III and IV of this notice, respectively.

II. Joint Technical Work Completed for This Final Rule

A. Introduction

In this section NHTSA and EPA discuss several aspects of the joint
technical analyses on which the two agencies collaborated. These
analyses are common to the development of each agency's final
standards. Specifically we discuss: the development of the vehicle
market forecast used by each agency for assessing costs, benefits, and
effects, the development of the attribute-based standard curve shapes,
the determination of the relative stringency between the car and truck
fleet standards, the technologies the agencies evaluated and their
costs and effectiveness, and the economic assumptions the agencies
included in their analyses. The Joint Technical Support Document (TSD)
discusses the agencies' joint technical work in more detail.

B. Developing the Future Fleet for Assessing Costs, Benefits, and
Effects

1. Why did the agencies establish a baseline and reference vehicle
fleet?
In order to calculate the impacts of the EPA and NHTSA regulations,
it is necessary to estimate the composition of the future vehicle fleet
absent these regulations, to provide a reference point relative to
which costs, benefits, and effects of the regulations are assessed. As
in the proposal, EPA and NHTSA have developed this comparison fleet in
two parts. The first step was to develop a baseline fleet based on
model year 2008 data. The second step was to project that fleet into
model years 2011-2016. This is called the reference fleet.

[[Page 25349]]

The third step was to modify that MY 2011-2016 reference fleet such
that it had sufficient technology to meet the MY 2011 CAFE standards.
This final version of the reference fleet is the light-duty fleet
estimated to exist in MY 2012-2016 in the absence of today's standards,
based on the assumption that manufacturers would continue to meet the
MY 2011 CAFE standards (or pay civil penalties allowed under EPCA \44\)
in the absence of further increases in the stringency of CAFE
standards. Each agency used this approach to develop a final reference
fleet to use in its modeling. All of the agencies' estimates of
emission reductions, fuel economy improvements, costs, and societal
impacts are developed in relation to the respective reference fleets.
---------------------------------------------------------------------------

\44\ That is, the manufacturers who have traditionally paid
fines under EPCA instead of complying with the CAFE standards were
``allowed,'' for purposes of the reference fleet, to reach only the
CAFE level at which paying fines became more cost-effective than
adding technology, even if that fell short of the MY 2011 standards.
---------------------------------------------------------------------------

EPA and NHTSA proposed a transparent approach to developing the
baseline and reference fleets, largely working from publicly available
data. This proposed approach differed from previous CAFE rules, which
relied on confidential manufacturers' product plan information to
develop the baseline. Most of the public comments to the NPRM
addressing this issue supported this methodology for developing the
inputs to the rule's analysis. Because the input sheets can be made
public, stakeholders can verify and check EPA's and NHTSA's modeling,
and perform their own analyses with these datasets. In this final
rulemaking, EPA and NHTSA are using an approach very similar to that
proposed, continuing to rely on publicly available data as the basis
for the baseline and reference fleets.
2. How did the agencies develop the baseline vehicle fleet?
At proposal, EPA and NHTSA developed a baseline fleet comprised of
model year 2008 data gathered from EPA's emission certification and
fuel economy database. MY 2008 was used as the basis for the baseline
vehicle fleet because it was the most recent model year for which a
complete set of data is publicly available. This remains the case.
Manufacturers are not required to submit final sales and mpg figures
for MY 2009 until April 2010,\45\ after the CAFE standard's mandated
promulgation date. Consequently, in this final rule, EPA and NHTSA made
no changes to the method or the results of the MY 2008 baseline fleet
used at proposal, except for some specific corrections to engineering
inputs for some vehicle models reflected in the market forecast input
to NHTSA's CAFE model. More details about how the agencies constructed
this baseline fleet can be found in Chapter 1.2 of the Joint TSD.
Corrections to engineering inputs for some vehicle models in the market
forecast input to NHTSA's CAFE model are discussed in Chapter 2 of the
Joint TSD.
---------------------------------------------------------------------------

\45\ 40 CFR 600.512-08, Model Year Report.
---------------------------------------------------------------------------

3. How did the agencies develop the projected MY 2011-2016 vehicle
fleet?
EPA and NHTSA have based the projection of total car and total
light truck sales for MYs 2011-2016 on projections made by the
Department of Energy's Energy Information Administration (EIA). EIA
publishes a mid-term projection of national energy use called the
Annual Energy Outlook (AEO). This projection utilizes a number of
technical and econometric models which are designed to reflect both
economic and regulatory conditions expected to exist in the future. In
support of its projection of fuel use by light-duty vehicles, EIA
projects sales of new cars and light trucks. In the proposal, the
agencies used the three reports published by EIA as part of the AEO
2009. We also stated that updated versions of these reports could be
used in the final rules should AEO timely issue a new version. EIA
published an early version of its AEO 2010 in December 2009, and the
agencies are making use of it in this final rulemaking. The differences
in projected sales in the 2009 report (used in the NPRM) and the early
2010 report are very small, so NHTSA and EPA have decided to simply
scale the NPRM volumes for cars and trucks (in the aggregate) to match
those in the 2010 report. We thus employ the sales projections from the
scaled updated 2009 Annual Energy Outlook, which is equivalent to AEO
2010 Early Release, for the final rule. The scaling factors for each
model year are presented in Chapter 1 of the Joint TSD for this final
rule.
The agencies recognize that AEO 2010 Early Release does include
some impacts of future projected increases in CAFE stringency. We have
closely examined the difference between AEO 2009 and AEO 2010 Early
Release and we believe the differences in total sales and the car/truck
split attributed to considerations of the standard in the final rule
are small.\46\
---------------------------------------------------------------------------

\46\ The agencies have also looked at the impact of the rule in
EIA's projection, and concluded that the impact was small. EPA and
NHTSA have evaluated the differences between the AEO 2010 (early
draft) and AEO 2009 and found little difference in the fleet
projections (or fuel prices). This analysis can be found in the memo
to the docket: Kahan, A. and Pickrell, D. Memo to Docket EPA-HQ-OAR-
2009-0472 and Docket NHTSA-2009-0059. ``Energy Information
Administration's Annual Energy Outlook 2009 and 2010.'' March 24,
2010.
---------------------------------------------------------------------------

In the AEO 2010 Early Release, EIA projects that total light-duty
vehicle sales will gradually recover from their currently depressed
levels by around 2013. In 2016, car sales are projected to be 9.4
million (57 percent) and truck sales are projected to be 7.1 million
(43 percent). Although the total level of sales of 16.5 million units
is similar to pre-2008 levels, the fraction of car sales is projected
to be higher than that existing in the 2000-2007 timeframe. This
projection reflects the impact of higher fuel prices, as well as EISA's
requirement that the new vehicle fleet average at least 35 mpg by MY
2020. The agencies note that AEO does not represent the fleet at a
level of detail sufficient to explicitly account for the
reclassification--promulgated as part of NHTSA's final rule for MY 2011
CAFE standards--of a number of 2-wheel drive sport utility vehicles
from the truck fleet to the car fleet for MYs 2011 and after. Sales
projections of cars and trucks for future model years can be found in
the Joint TSD for these final rules.
In addition to a shift towards more car sales, sales of segments
within both the car and truck markets have been changing and are
expected to continue to change. Manufacturers are introducing more
crossover models which offer much of the utility of SUVs but use more
car-like designs. The AEO 2010 report does not, however, distinguish
such changes within the car and truck classes. In order to reflect
these changes in fleet makeup, EPA and NHTSA considered several other
available forecasts. EPA purchased and shared with NHTSA forecasts from
two well-known industry analysts, CSM Worldwide (CSM), and J.D. Powers.
NHTSA and EPA decided to use the forecast from CSM, modified as
described below, for several reasons presented in the NPRM preamble
\47\ and draft Joint TSD. The changes between company market share and
industry market segments were most significant from 2011-2014, while
for 2014-2015 the changes were relatively small. Noting this, and
lacking a credible forecast of company and segment shares after 2015,
the agencies assumed 2016 market share and market segments to be the
same as for 2015.
---------------------------------------------------------------------------

\47\ See, e.g., 74 FR 49484.

---------------------------------------------------------------------------

[[Page 25350]]

CSM Worldwide provides quarterly sales forecasts for the automotive
industry. In the NPRM, the agencies identified a concern with the 2nd
quarter CSM forecast that was used as a basis for the projection. CSM
projections at that time were based on an industry that was going
through a significant financial transition, and as a result the market
share forecasts for some companies were impacted in surprising ways. As
the industry's situation has settled somewhat over the past year, the
4th quarter projection appears to address this issue--for example, it
shows nearly a two-fold increase in sales for Chrysler compared to
significant loss of market share shown for Chrysler in the 2nd quarter
projection. Additionally, some commenters, such as GM, recognized that
the fleet appeared to include an unusually high number of large pickup
trucks.\48\ In fact, the agencies discovered (independently of the
comments) that CSM's standard forecast included all vehicles below
14,000 GVWR, including class 2b and 3 heavy duty vehicles, which are
not regulated by this final rule.\49\ The commenters were thus correct
that light duty reference fleet projections at proposal had more full
size trucks and vans due to the mistaken inclusion of the heavy duty
versions of those vehicles. The agencies requested a separate data
forecast from CSM that filtered their 4th quarter projection to exclude
these heavy duty vehicles. The agencies then used this filtered 4th
quarter forecast for the final rule. A detailed comparison of the
market by manufacturer can be found in the final TSD. For the public's
reference, copies of the 2nd, 3rd, and 4th quarter CSM forecasts have
been placed in the docket for this rulemaking.\50\
---------------------------------------------------------------------------

\48\ GM argued that the unusually large volume of large pickups
led to higher overall requirements for those vehicles. As discussed
below, the agencies' analysis for the final rule corrects the number
of large pickups. With this correction and other updates to the
agencies' market forecast and other analytical inputs, the target
functions defining the final standards (and achieving the average
required performance levels defining the national program) are very
similar to those from the NPRM, especially for light trucks, as
illustrated below in Figures II.C-7 and II.C-8.
\49\ These include the Ford F-250 & F-350, Econoline E-250, & E-
350; Chevy Express, Silverado 2500, & 3500; GMC Savana, Dodge 2500,
& 3500; among others.
\50\ The CSM Sales Forecast Excel file (``CSM North America
Sales Forecasts 2Q09 3Q09 4Q09 for the Docket'') is available in the
docket (Docket EPA-HQ-OAR-2009-0472).
---------------------------------------------------------------------------

We then projected the CSM forecasts for relative sales of cars and
trucks by manufacturer and by market segment onto the total sales
estimates of AEO 2010. Tables II.B.3-1 and II.B.3-2 show the resulting
projections for the reference 2016 model year and compare these to
actual sales that occurred in baseline 2008 model year. Both tables
show sales using the traditional definition of cars and light trucks.

Table II.B.3-1--Annual Sales of Light-Duty Vehicles by Manufacturer in 2008 and Estimated for 2016
----------------------------------------------------------------------------------------------------------------
Cars Light trucks Total
-----------------------------------------------------------------------------
2008 MY 2016 MY 2008 MY 2016 MY 2008 MY 2016 MY
----------------------------------------------------------------------------------------------------------------
BMW............................... 291,796 424,923 61,324 171,560 353,120 596,482
Chrysler.......................... 537,808 340,908 1,119,397 525,128 1,657,205 866,037
Daimler........................... 208,052 272,252 79,135 126,880 287,187 399,133
Ford.............................. 709,583 1,118,727 1,158,805 1,363,256 1,868,388 2,481,983
General Motors.................... 1,370,280 1,283,937 1,749,227 1,585,828 3,119,507 2,869,766
Honda............................. 899,498 811,214 612,281 671,437 1,511,779 1,482,651
Hyundai........................... 270,293 401,372 120,734 211,996 391,027 613,368
Kia............................... 145,863 455,643 135,589 210,717 281,452 666,360
Mazda............................. 191,326 350,055 111,220 144,992 302,546 495,047
Mitsubishi........................ 76,701 49,914 24,028 88,754 100,729 138,668
Porsche........................... 18,909 33,471 18,797 16,749 37,706 50,220
Nissan............................ 653,121 876,677 370,294 457,114 1,023,415 1,333,790
Subaru............................ 149,370 230,705 49,211 95,054 198,581 325,760
Suzuki............................ 68,720 97,466 45,938 26,108 114,658 123,574
Tata.............................. 9,596 65,806 55,584 42,695 65,180 108,501
Toyota............................ 1,143,696 2,069,283 1,067,804 1,249,719 2,211,500 3,319,002
Volkswagen........................ 290,385 586,011 26,999 124,703 317,384 710,011
-----------------------------------------------------------------------------
Total......................... 7,034,997 9,468,365 6,806,367 7,112,689 13,841,364 16,580,353
----------------------------------------------------------------------------------------------------------------

Table II.B.3-2--Annual Sales of Light-Duty Vehicles by Market Segment in 2008 and Estimated for 2016
----------------------------------------------------------------------------------------------------------------
Cars Light trucks
----------------------------------------------------------------------------------------------------------------
2008 MY 2016 MY 2008 MY 2016 MY
----------------------------------------------------------------------------------------------------------------
Full-Size Car................. 829,896 530,945 Full-Size Pickup 1,331,989 1,379,036
Luxury Car.................... 1,048,341 1,548,242 Mid-Size Pickup. 452,013 332,082
Mid-Size Car.................. 2,166,849 2,550,561 Full-Size Van... 33,384 65,650
Mini Car...................... 617,902 1,565,373 Mid-Size Van.... 719,529 839,194
Small Car..................... 1,912,736 2,503,566 Mid-Size MAV *.. 110,353 116,077
Specialty Car................. 459,273 769,679 Small MAV....... 231,265 62,514
Full-Size SUV *. 559,160 232,619
Mid-Size SUV.... 436,080 162,502
Small SUV....... 196,424 108,858
Full-Size CUV *. 264,717 260,662
Mid-Size CUV.... 923,165 1,372,200
Small CUV....... 1,548,288 2,181,296
---------------------------------------------------------------------------------

[[Page 25351]]

Total Sales **............ 7,034,997 9,468,365 ................ 6,806,367 7,079,323
----------------------------------------------------------------------------------------------------------------
* MAV--Multi-Activity Vehicle, SUV--Sport Utility Vehicle, CUV--Crossover Utility Vehicle.
** Total Sales are based on the classic Car/Truck definition.

Determining which traditionally-defined trucks will be defined as
cars for purposes of this final rule using the revised definition
established by NHTSA for MYs 2011 and beyond requires more detailed
information about each vehicle model. This is described in greater
detail in Chapter 1 of the final TSD.
The forecasts obtained from CSM provided estimates of car and truck
sales by segment and by manufacturer, but not by manufacturer for each
market segment. Therefore, NHTSA and EPA needed other information on
which to base these more detailed projected market splits. For this
task, the agencies used as a starting point each manufacturer's sales
by market segment from model year 2008, which is the baseline fleet.
Because of the larger number of segments in the truck market, the
agencies used slightly different methodologies for cars and trucks.
The first step for both cars and trucks was to break down each
manufacturer's 2008 sales according to the market segment definitions
used by CSM. For example, the agencies found that Ford's \51\ cars
sales in 2008 were broken down as shown in Table II.B.3-3:
---------------------------------------------------------------------------

\51\ Note: In the NPRM, Ford's 2008 sales per segment, and the
total number of cars was different than shown here. The change in
values is due to a correction of vehicle segments for some of Ford's
vehicles.

Table II.B.3-3--Breakdown of Ford's 2008 Car Sales
------------------------------------------------------------------------

------------------------------------------------------------------------
Full-size cars.......................... 160,857 units.
Mid-size Cars........................... 170,399 units.
Small/Compact Cars...................... 180,249 units.
Subcompact/Mini Cars.................... None.
Luxury cars............................. 87,272 units.
Specialty cars.......................... 110,805 units.
------------------------------------------------------------------------

EPA and NHTSA then adjusted each manufacturer's sales of each of
its car segments (and truck segments, separately) so that the
manufacturer's total sales of cars (and trucks) matched the total
estimated for each future model year based on AEO and CSM forecasts.
For example, as indicated in Table II.B.3-1, Ford's total car sales in
2008 were 709,583 units, while the agencies project that they will
increase to 1,113,333 units by 2016. This represents an increase of
56.9 percent. Thus, the agencies increased the 2008 sales of each Ford
car segment by 56.9 percent. This produced estimates of future sales
which matched total car and truck sales per AEO and the manufacturer
breakdowns per CSM. However, the sales splits by market segment would
not necessarily match those of CSM (shown for 2016 in Table II.B.3-2).
In order to adjust the market segment mix for cars, the agencies
first adjusted sales of luxury, specialty and other cars. Since the
total sales of cars for each manufacturer were already set, any changes
in the sales of one car segment had to be compensated by the opposite
change in another segment. For the luxury, specialty and other car
segments, it is not clear how changes in sales would be compensated.
For example, if luxury car sales decreased, would sales of full-size
cars increase, mid-size cars, and so on? The agencies have assumed that
any changes in the sales of cars within these three segments were
compensated for by proportional changes in the sales of the other four
car segments. For example, for 2016, the figures in Table II.B.3-2
indicate that luxury car sales in 2016 are 1,548,242 units. Luxury car
sales are 1,048,341 units in 2008. However, after adjusting 2008 car
sales by the change in total car sales for 2016 projected by EIA and a
change in manufacturer market share per CSM, luxury car sales decreased
to 1,523,171 units. Thus, overall for 2016, luxury car sales had to
increase by 25,071 units or 6 percent. The agencies accordingly
increased the luxury car sales by each manufacturer by this percentage.
The absolute decrease in luxury car sales was spread across sales of
full-size, mid-size, compact and subcompact cars in proportion to each
manufacturer's sales in these segments in 2008. The same adjustment
process was used for specialty cars and the ``other cars'' segment
defined by CSM.
The agencies used a slightly different approach to adjust for
changing sales of the remaining four car segments. Starting with full-
size cars, the agencies again determined the overall percentage change
that needed to occur in future year full-size car sales after 1)
adjusting for total sales per AEO 2010, 2) adjusting for manufacturer
sales mix per CSM and 3) adjusting the luxury, specialty and other car
segments, in order to meet the segment sales mix per CSM. Sales of each
manufacturer's large cars were adjusted by this percentage. However,
instead of spreading this change over the remaining three segments, the
agencies assigned the entire change to mid-size vehicles. The agencies
did so because the CSM data followed the trend of increasing volumes of
smaller cars while reducing volumes of larger cars. If a consumer had
previously purchased a full-size car, we thought it unlikely that their
next purchase would decrease by two size categories, down to a
subcompact. It seemed more reasonable to project that they would drop
one vehicle size category smaller. Thus, the change in each
manufacturer's sales of full-size cars was matched by an opposite
change (in absolute units sold) in mid-size cars.
The same process was then applied to mid-size cars, with the change
in mid-size car sales being matched by an opposite change in compact
car sales. This process was repeated one more time for compact car
sales, with changes in sales in this segment being matched by the
opposite change in the sales of subcompacts. The overall result was a
projection of car sales for model years 2012-2016--the reference
fleet--which matched the total sales projections of the AEO forecast
and the manufacturer and segment splits of the CSM forecast. These
sales splits can be found in Chapter 1 of the Joint TSD for this final
rule.
As mentioned above, the agencies applied a slightly different
process to truck sales, because the agencies could not confidently
project how the change in sales from one segment preferentially went to
or came from another particular segment. Some trend from larger
vehicles to smaller vehicles would have been possible. However, the CSM
forecasts indicated large changes in total sport utility vehicle,
multi-activity vehicle and cross-over sales which could not be
connected. Thus, the

[[Page 25352]]

agencies applied an iterative, but straightforward process for
adjusting 2008 truck sales to match the AEO and CSM forecasts.
The first three steps were exactly the same as for cars. EPA and
NHTSA broke down each manufacturer's truck sales into the truck
segments as defined by CSM. The agencies then adjusted all
manufacturers' truck segment sales by the same factor so that total
truck sales in each model year matched AEO projections for truck sales
by model year. The agencies then adjusted each manufacturer's truck
sales by segment proportionally so that each manufacturer's percentage
of total truck sales matched that forecast by CSM. This again left the
need to adjust truck sales by segment to match the CSM forecast for
each model year.
In the fourth step, the agencies adjusted the sales of each truck
segment by a common factor so that total sales for that segment matched
the combination of the AEO and CSM forecasts. For example, projected
sales of large pickups across all manufacturers were 1,286,184 units in
2016 after adjusting total sales to match AEO's forecast and adjusting
each manufacturer's truck sales to match CSM's forecast for the
breakdown of sales by manufacturer. Applying CSM's forecast of the
large pickup segment of truck sales to AEO's total sales forecast
indicated total large pickup sales of 1,379,036 units. Thus, we
increased each manufacturer's sales of large pickups by 7 percent.\52\
The agencies applied the same type of adjustment to all the other truck
segments at the same time. The result was a set of sales projections
which matched AEO's total truck sales projection and CSM's market
segment forecast. However, after this step, sales by manufacturer no
longer met CSM's forecast. Thus, we repeated step three and adjusted
each manufacturer's truck sales so that they met CSM's forecast. The
sales of each truck segment (by manufacturer) were adjusted by the same
factor. The resulting sales projection matched AEO's total truck sales
projection and CSM's manufacturer forecast, but sales by market segment
no longer met CSM's forecast. However, the difference between the sales
projections after this fifth step was closer to CSM's market segment
forecast than it was after step three. In other words, the sales
projection was converging to the desired result. The agencies repeated
these adjustments, matching manufacturer sales mix in one step and then
market segment in the next a total of 19 times. At this point, we were
able to match the market segment splits exactly and the manufacturer
splits were within 0.1 percent of our goal, which is well within the
needs of this analysis.
---------------------------------------------------------------------------

\52\ Note: In the NPRM this example showed 29 percent instead of
7 percent. The significant decrease was due to using the filtered
4th quarter CSM forecast. Commenters, such as GM, had commented that
we had too many full-size trucks and vans, and this change addresses
their comment.
---------------------------------------------------------------------------

The next step in developing the reference fleets was to
characterize the vehicles within each manufacturer-segment combination.
In large part, this was based on the characterization of the specific
vehicle models sold in 2008--i.e., the vehicles comprising the baseline
fleet. EPA and NHTSA chose to base our estimates of detailed vehicle
characteristics on 2008 sales for several reasons. One, these vehicle
characteristics are not confidential and can thus be published here for
careful review by interested parties. Two, because it is constructed
beginning with actual sales data, this vehicle fleet is limited to
vehicle models known to satisfy consumer demands in light of price,
utility, performance, safety, and other vehicle attributes.
As noted above, the agencies gathered most of the information about
the 2008 baseline vehicle fleet from EPA's emission certification and
fuel economy database. The data obtained from this source included
vehicle production volume, fuel economy, engine size, number of engine
cylinders, transmission type, fuel type, etc. EPA's certification
database does not include a detailed description of the types of fuel
economy-improving/CO2-reducing technologies considered in
this final rule. Thus, the agencies augmented this description with
publicly available data which includes more complete technology
descriptions from Ward's Automotive Group.\53\ In a few instances when
required vehicle information (such as vehicle footprint) was not
available from these two sources, the agencies obtained this
information from publicly accessible Internet sites such as
Motortrend.com and Edmunds.com.\54\
---------------------------------------------------------------------------

\53\ Note that WardsAuto.com is a fee-based service, but all
information is public to subscribers.
\54\ Motortrend.com and Edmunds.com are free, no-fee Internet
sites.
---------------------------------------------------------------------------

The projections of future car and truck sales described above apply
to each manufacturer's sales by market segment. The EPA emissions
certification sales data are available at a much finer level of detail,
essentially vehicle configuration. As mentioned above, the agencies
placed each vehicle in the EPA certification database into one of the
CSM market segments. The agencies then totaled the sales by each
manufacturer for each market segment. If the combination of AEO and CSM
forecasts indicated an increase in a given manufacturer's sales of a
particular market segment, then the sales of all the individual vehicle
configurations were adjusted by the same factor. For example, if the
Prius represented 30 percent of Toyota's sales of compact cars in 2008
and Toyota's sales of compact cars in 2016 was projected to double by
2016, then the sales of the Prius were doubled, and the Prius sales in
2016 remained 30 percent of Toyota's compact car sales.
The projection of average footprint for both cars and trucks
remained virtually constant over the years covered by the final
rulemaking. This occurrence is strictly a result of the CSM
projections. There are a number of trends that occur in the CSM
projections that caused the average footprint to remain constant.
First, as the number of subcompacts increases, so do the number of 2-
wheel drive crossover vehicles (that are regulated as cars). Second,
truck volumes have many segment changes during the rulemaking time
frame. There is no specific footprint related trend in any segment that
can be linked to the unchanging footprint, but there is a trend that
non-pickups' volumes will move from truck segments that are ladder
frame to those that are unibody-type vehicles. A table of the footprint
projections is available in the TSD as well as further discussion on
this topic.
4. How was the development of the baseline and reference fleets for
this Final Rule different from NHTSA's historical approach?
NHTSA has historically based its analysis of potential new CAFE
standards on detailed product plans the agency has requested from
manufacturers planning to produce light vehicles for sale in the United
States. Although the agency has not attempted to compel manufacturers
to submit such information, most major manufacturers and some smaller
manufacturers have voluntarily provided it when requested.
The proposal discusses many of the advantages and disadvantages of
the market forecast approach used by the agencies, including the
agencies' interest in examining product plans as a check on the
reference fleet developed by the agencies for this rulemaking. One of
the primary reasons for the request for data in 2009 was to obtain
permission from the manufacturers to make public their product plan
information for model years 2010 and 2011. There are a number of
reasons that this could be advantageous in the development of a
reference fleet. First,

[[Page 25353]]

some known changes to the fleet may not be captured by the approach of
solely using publicly available information. For example, the agencies'
current market forecast includes some vehicles for which manufacturers
have announced plans for elimination or drastic production cuts such as
the Chevrolet Trailblazer, the Chrysler PT Cruiser, the Chrysler
Pacifica, the Dodge Magnum, the Ford Crown Victoria, the Mercury Sable,
the Pontiac Grand Prix, the Pontiac G5 and the Saturn Vue. These
vehicle models appear explicitly in market inputs to NHTSA's analysis,
and are among those vehicle models included in the aggregated vehicle
types appearing in market inputs to EPA's analysis. However, although
the agencies recognize that these specific vehicles will be
discontinued, we continue to include them in the market forecast
because they are useful as a surrogate for successor vehicles that may
appear in the rulemaking time frame to replace the discontinued
vehicles in that market segment.\55\
---------------------------------------------------------------------------

\55\ An example of this is in the GM Pontiac line, which is in
the process of being phased out during the course of this
rulemaking. GM has similar vehicles within their other brands (like
Chevy) that will ``presumably'' pick up the loss in Pontiac share.
We model this simply by leaving the Pontiac brand in.
---------------------------------------------------------------------------

Second, the agencies' market forecast does not include some
forthcoming vehicle models, such as the Chevrolet Volt, the Ford Fiesta
and several publicly announced electric vehicles, including the
announcements from Nissan regarding the Leaf. Nor does it include
several MY 2009 or 2010 vehicles, such as the Honda Insight, the
Hyundai Genesis and the Toyota Venza, as our starting point for
defining specific vehicle models in the reference fleet was Model Year
2008. Additionally, the market forecast does not account for publicly
announced technology introductions, such as Ford's EcoBoost system,
whose product plans specify which vehicles and how many are planned to
have this technology. Chrysler Group LLC has announced plans to offer
small- and medium-sized cars using Fiat powertrains. Were the agencies
to rely on manufacturers' product plans (that were submitted), the
market forecast would account for not only these specific examples, but
also for similar examples that have not yet been announced publicly.
Some commenters, such as CBD and NESCAUM, suggested that the
agencies' omission of known future vehicles and technologies in the
reference fleet causes inaccuracies, which CBD further suggested could
lead the agencies to set lower standards. On the other hand, CARB
commented that ``the likely impact of this omission is minor.'' Because
the agencies' analysis examines the costs and benefits of progressively
adding technology to manufacturers' fleets, the omission of future
vehicles and technologies primarily affects how much additional
technology (and, therefore, how much incremental cost and benefit) is
available relative to the point at which the agencies' examination of
potential new standards begins. Thus, in fact, the omission only
reflects the reference fleet, rather than the agencies' conclusions
regarding how stringent the standards should be. This is discussed
further below. The agencies believe the above-mentioned comments by
CBD, NESCAUM, and others are based on a misunderstanding of the
agencies' approach to analyzing potential increases in regulatory
stringency. The agencies also note that manufacturers do not always use
technology solely to increase fuel economy, and that use of technology
to increase vehicles' acceleration performance or utility would
probably make that technology unavailable toward more stringent
standards. Considering the incremental nature of the agencies'
analysis, and the counterbalancing aspects of potentially omitted
technology in the reference fleet, the agencies believe their
determination of the stringency of new standards has not been impacted
by any such omissions.
Moreover, EPA and NHTSA believe that not including such vehicles
after MY 2008 does not significantly impact our estimates of the
technology required to comply with the standards. If included, these
vehicles could increase the extent to which manufacturers are, in the
reference case, expected to over-comply with the MY 2011 CAFE
standards, and could thereby make the new standards appear to cost less
and yield less benefit relative to the reference case. However, in the
agencies' judgment, production of the most advanced technology
vehicles, such as the Chevy Volt or the Nissan Leaf (for example), will
most likely be too limited during MY 2011 through MY 2016 to
significantly impact manufacturers' compliance positions. While we are
projecting the characteristics of the future fleet by extrapolating
from the MY 2008 fleet, the primary difference between the future fleet
and the 2008 fleet in the same vehicle segment is the use of additional
CO2-reducing and fuel-saving technologies. Both the NHTSA
and EPA models add such technologies to evaluate means of complying
with the standards, and the costs of doing so. Thus, our future
projections of the vehicle fleet generally shift vehicle designs
towards those more likely to be typical of newer vehicles. Compared to
using product plans that show continued fuel economy increases planned
based on expectations that CAFE standards will continue to increase,
this approach helps to clarify the costs and benefits of the new
standards, as the costs and benefits of all fuel economy improvements
beyond those required by the MY 2011 CAFE standards are being assigned
to the final rules. In some cases, the ``actual'' (vs. projected or
``modeled'') new vehicles being introduced into the market by
manufacturers are done so in anticipation of this rulemaking. On the
other hand, manufacturers may plan to continue using technologies to
improve vehicle performance and/or utility, not just fuel economy. Our
approach prevents some of these actual technological improvements and
their associated cost and fuel economy improvements from being assumed
in the reference fleet. Thus, the added technology will not be
considered to be free (or having no benefits) for the purposes of this
rule.
In this regard, the agencies further note that manufacturer
announcements regarding forward models (or future vehicle models) need
not be accepted automatically. Manufacturers tend to limit accurate
production intent information in these releases for reasons such as:
(a) Competitors will closely examine their information for data in
their product planning decisions; (b) the press coverage of forward
model announcements is not uniform, meaning highly anticipated models
have more coverage and materials than models that may be less exciting
to the public and consistency and uniformity cannot be ensured with the
usage of press information; and (c) these market projections are
subject to change (sometimes significant), and manufacturers may not
want to give the appearance of being indecisive, or under/over-
confident to their shareholders and the public with premature release
of information.
NHTSA has evaluated the use of public manufacturer forward model
press information to update the vehicle fleet inputs to the baseline
and reference fleet. The challenges in this approach are evidenced by
the continuous stream of manufacturer press releases throughout a
defined rulemaking period. Manufacturers' press releases suffer from
the same types of inaccuracies that many commenters believe can affect
product plans.

[[Page 25354]]

Manufacturers can often be overly optimistic in their press releases,
both on projected date of release of new models and on sales volumes.
More generally and more critically, as discussed in the proposal
and as endorsed by many of the public comments, there are several
advantages to the approach used by the agencies in this final rule.
Most importantly, today's market forecast is much more transparent. The
information sources used to develop today's market forecast are all
either in the public domain or available commercially. Another
significant advantage of today's market forecast is the agencies'
ability to assess more fully the incremental costs and benefits of the
proposed standards. In addition, by developing baseline and reference
fleets from common sources, the agencies have been able to avoid some
errors--perhaps related to interpretation of requests--that have been
observed in past responses to NHTSA's requests. An additional advantage
of the approach used for this rule is a consistent projection of the
change in fuel economy and CO2 emissions across the various
vehicles from the application of new technology. With the approach used
for this final rule, the baseline market data comes from actual
vehicles (on the road today) which have actual fuel economy test data
(in contrast to manufacturer estimates of future product fuel
economy)--so there is no question what is the basis for the fuel
economy or CO2 performance of the baseline market data as it
is.
5. How does manufacturer product plan data factor into the baseline
used in this Final Rule?
In the spring and fall of 2009, many manufacturers submitted
product plans in response to NHTSA's recent requests that they do so.
NHTSA and EPA both have access to these plans, and both agencies have
reviewed them in detail. A small amount of product plan data was used
in the development of the baseline. The specific pieces of data are:
Wheelbase.
Track Width Front.
Track Width Rear.
EPS (Electric Power Steering).
ROLL (Reduced Rolling Resistance).
LUB (Advance Lubrication i.e. low weight oil).
IACC (Improved Electrical Accessories).
Curb Weight.
GVWR (Gross Vehicle Weight Rating).
The track widths, wheelbase, curb weight, and GVWR for vehicles
could have been looked up on the Internet (159 were), but were taken
from the product plans when available for convenience. To ensure
accuracy, a sample from each product plan was used as a check against
the numbers available from Motortrend.com. These numbers will be
published in the baseline file since they can be easily looked up on
the internet. On the other hand, EPS, ROLL, LUB, and IACC are difficult
to determine without using manufacturer's product plans. These items
will not be published in the baseline file, but the data has been
aggregated into the agencies' baseline in the technology effectiveness
and cost effectiveness for each vehicle in a way that allows the
baseline for the model to be published without revealing the
manufacturer's data.
Also, some technical information that manufacturers have provided
in product plans regarding specific vehicle models is, at least insofar
as NHTSA and EPA have been able to determine, not available from public
or commercial sources. While such gaps do not bear significantly on the
agencies' analysis, the diversity of pickup configurations necessitated
utilizing a sales-weighted average footprint value \56\ for many
manufacturers' pickups. Since our modeling only utilizes footprint in
order to estimate each manufacturer's CO2 or fuel economy
standard and all the other vehicle characteristics are available for
each pickup configuration, this approximation has no practical impact
on the projected technology or cost associated with compliance with the
various standards evaluated. The only impact which could arise would be
if the relative sales of the various pickup configurations changed, or
if the agencies were to explore standards with a different shape. This
would necessitate recalculating the average footprint value in order to
maintain accuracy.
---------------------------------------------------------------------------

\56\ A full-size pickup might be offered with various
combinations of cab style (e.g., regular, extended, crew) and box
length (e.g., 5\1/2\', 6\1/2\', 8') and, therefore, multiple
footprint sizes. CAFE compliance data for MY 2008 data does not
contain footprint information, and does not contain information that
can be used to reliably identify which pickup entries correspond to
footprint values estimable from public or commercial sources.
Therefore, the agencies have used the known production levels of
average values to represent all variants of a given pickup line
(e.g., all variants of the F-150 and the Sierra/Silverado) in order
to calculate the sales-weighted average footprint value for each
pickup family. Again, this has no impact on the results of our
modeling effort, although it would require re-estimation if we were
to examine light truck standards of a different shape. In the
extreme, one single footprint value could be used for every vehicle
sold by a single manufacturer as long as the fuel economy standard
associated with this footprint value represented the sales-weighted,
harmonic average of the fuel economy standards associated with each
vehicle's footprint values.
---------------------------------------------------------------------------

Additionally, as discussed in the NPRM, in an effort to update the
2008 baseline to account for the expected changes in the fleet in the
near-term model years 2009-2011 described above, NHTSA requested
permission from the manufacturers to make this limited product plan
information public. Unfortunately, virtually no manufacturers agreed to
allow the use of their data after 2009 model year. A few manufacturers,
such as GM and Ford, stated we could use their 2009 product plan data
after the end of production (December 31), but this would not have
afforded us sufficient time to do the analysis for the final rule.
Since the agencies were unable to obtain consistent updates, the
baseline and reference fleets were not updated beyond 2008 model year
for the final rule. The 2008 baseline fleet and projections were
instead updated using the latest AEO and CSM data as discussed earlier.
NHTSA and EPA recognize that the approach applied for the current
rule gives transparency and openness of the vehicle market forecast
high priority, and accommodates minor inaccuracies that may be
introduced by not accounting for future product mix changes anticipated
in manufacturers' confidential product plans. For any future fleet
analysis that the agencies are required to perform, NHTSA and EPA plan
to request that manufacturers submit product plans and allow some
public release of information. In performing this analysis, the
agencies plan to reexamine potential tradeoffs between transparency and
technical reasonableness, and to explain resultant choices.

C. Development of Attribute-Based Curve Shapes

In the NPRM, NHTSA and EPA proposed to set attribute-based CAFE and
CO2 standards that are defined by a mathematical function
for MYs 2012-2016 passenger cars and light trucks. EPCA, as amended by
EISA, expressly requires that CAFE standards for passenger cars and
light trucks be based on one or more vehicle attributes related to fuel
economy, and be expressed in the form of a mathematical function.\57\
The CAA has no such requirement, though in past rules, EPA has relied
on both universal and attribute-based standards (e.g., for nonroad
engines, EPA uses the attribute of horsepower). However, given the
advantages of using attribute-based standards and given the

[[Page 25355]]

goal of coordinating and harmonizing CO2 standards
promulgated under the CAA and CAFE standards promulgated under EPCA,
EPA also proposed to issue standards that are attribute-based and
defined by mathematical functions. There was consensus in the public
comments that EPA should develop attribute-based CO2
standards.
---------------------------------------------------------------------------

\57\ 49 U.S.C. 32902(a)(3)(A).
---------------------------------------------------------------------------

Comments received in response to the agencies' decision to base
standards on vehicle footprint were largely supportive. Several
commenters (BMW, NADA, NESCAUM) expressed support for attribute-based
(as opposed to flat or universal) standards generally, and agreed with
EPA's decision to harmonize with NHTSA in this respect. Many commenters
(Aluminum Association, BMW, ICCT, NESCAUM, NY DEC, Schade, Toyota) also
supported the agencies' decision to continue setting CAFE standards,
and begin setting GHG standards, on the basis of vehicle footprint,
although one commenter (NJ DEP) opposed the use of footprint due to
concern that it encourages manufacturers to upsize vehicles and
undercut the gains of the standard. Of the commenters supporting the
use of footprint, several focused on the benefits of harmonization--
both between EPA and NHTSA, and between the U.S. and the rest of the
world. BMW commented, for example, that many other countries use
weight-based standards rather than footprint-based. While BMW did not
object to NHTSA's and EPA's use of footprint-based standards, it
emphasized the impact of this non-harmonization on manufacturers who
sell vehicles globally, and asked the agencies to consider these
effects. NADA supported the use of footprint, but cautioned that the
agencies must be careful in setting the footprint curve for light
trucks to ensure that manufacturers can continue to provide
functionality like 4WD and towing/hauling capacity.
Some commenters requested that the agencies consider other or more
attributes in addition to footprint, largely reiterating comments
submitted to the MYs 2011-2015 CAFE NPRM. Cummins supported the
agencies using a secondary attribute to account for towing and hauling
capacity in large trucks, for example, while Ferrari asked the agencies
to consider a multi-attribute approach incorporating curb weight,
maximum engine power or torque, and/or engine displacement, as it had
requested in the previous round of CAFE rulemaking. An individual, Mr.
Kenneth Johnson, commented that weight-based standards would be
preferable to footprint-based ones, because weight correlates better
with fuel economy than footprint, because the use of footprint does not
necessarily guarantee safety the way the agencies say it does, and
because weight-based standards would be fairer to manufacturers.
In response, EPA and NHTSA continue to believe that the benefits of
footprint-attribute-based standards outweigh any potential drawbacks
raised by commenters, and that harmonization between the two agencies
should be the overriding goal on this issue. As discussed by NHTSA in
the MY 2011 CAFE final rule,\58\ the agencies believe that the
possibility of gaming is lowest with footprint-based standards, as
opposed to weight-based or multi-attribute-based standards.
Specifically, standards that incorporate weight, torque, power, towing
capability, and/or off-road capability in addition to footprint would
not only be significantly more complex, but by providing degrees of
freedom with respect to more easily-adjusted attributes, they would
make it less certain that the future fleet would actually achieve the
average fuel economy and CO2 levels projected by the
agencies. The agencies recognize that based on economic and consumer
demand factors that are external to this rule, the distribution of
footprints in the future may be different (either smaller or larger)
than what is projected in this rule. However, the agencies continue to
believe that there will not be significant shifts in this distribution
as a direct consequence of this rule. The agencies are therefore
finalizing MYs 2012-2016 CAFE and GHG standards based on footprint.
---------------------------------------------------------------------------

\58\ See 74 FR 14359 (Mar. 30, 2009).
---------------------------------------------------------------------------

The agencies also recognize that there could be benefits for a
number of manufacturers if there was greater international
harmonization of fuel economy and GHG standards, but this is largely a
question of how stringent standards are and how they are enforced. It
is entirely possible that footprint-based and weight-based systems can
coexist internationally and not present an undue burden for
manufacturers if they are carefully crafted. Different countries or
regions may find different attributes appropriate for basing standards,
depending on the particular challenges they face--from fuel prices, to
family size and land use, to safety concerns, to fleet composition and
consumer preference, to other environmental challenges besides climate
change. The agencies anticipate working more closely with other
countries and regions in the future to consider how to mitigate these
issues in a way that least burdens manufacturers while respecting each
country's need to meet its own particular challenges.
Under an attribute-based standard, every vehicle model has a
performance target (fuel economy and CO2 emissions for CAFE
and CO2 emissions standards, respectively), the level of
which depends on the vehicle's attribute (for the proposal, footprint).
The manufacturers' fleet average performance is determined by the
production-weighted \59\ average (for CAFE, harmonic average) of those
targets. NHTSA and EPA are promulgating CAFE and CO2
emissions standards defined by constrained linear functions and,
equivalently, piecewise linear functions.\60\ As a possible option for
future rulemakings, the constrained linear form was introduced by NHTSA
in the 2007 NPRM proposing CAFE standards for MY 2011-2015. Described
mathematically, the proposed constrained linear function was defined
according to the following formula: \61\
---------------------------------------------------------------------------

\59\ Production for sale in the United States.
\60\ The equations are equivalent but are specified differently
due to differences in the agencies' respective models.
\61\ This function is linear in fuel consumption but not in fuel
economy.
[GRAPHIC] [TIFF OMITTED] TR07MY10.004

---------------------------------------------------------------------------
Where

TARGET = the fuel economy target (in mpg) applicable to vehicles of
a given footprint (FOOTPRINT, in square feet),
a = the function's upper limit (in mpg),
b = the function's lower limit (in mpg),

[[Page 25356]]

c = the slope (in gpm per square foot) of the sloped portion of the
function,
d = the intercept (in gpm) of the sloped portion of the function
(that is, the value the sloped portion would take if extended to a
footprint of 0 square feet, and the MIN and MAX functions take the
minimum and maximum, respectively, of the included values; for
example, MIN(1,2) = 1, MAX(1,2) = 2, and
MIN[MAX(1,2),3)]=2.

Because the format is linear on a gallons-per-mile basis, not on a
miles-per-gallon basis, it is plotted as fuel consumption below.
Graphically, the constrained linear form appears as shown in Figure
II.C-1.
BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TR07MY10.005

[[Page 25357]]

The specific form and stringency for each fleet (passenger car and
light trucks) and model year are defined through specific values for
the four coefficients shown above.
EPA proposed the equivalent equation below for assigning
CO2 targets to an individual vehicle's footprint value.
Although the general model of the equation is the same for each vehicle
category and each year, the parameters of the equation differ for cars
and trucks and for each model year. Described mathematically, EPA's
proposed piecewise linear function was as follows:

Target = a, if x <= l
Target = cx + d, if l < x <= h
Target = b, if x > h

In the constrained linear form similar in form to the fuel economy
equation above, this equation takes the simplified form:

Target = MIN [ MAX (c * x + d, a), b]

Where

Target = the CO2 target value for a given footprint (in
g/mi)
a = the minimum target value (in g/mi CO2) \62\
---------------------------------------------------------------------------

\62\ These a, b, d coefficients differ from the a, b, d
coefficients in the constrained linear fuel economy equation
primarily by a factor of 8887 (plus an additive factor for air
conditioning).
---------------------------------------------------------------------------

b = the maximum target value (in g/mi CO2)
c = the slope of the linear function (in g/mi per sq ft
CO2)
d = is the intercept or zero-offset for the line (in g/mi
CO2)
x = footprint of the vehicle model (in square feet, rounded to the
nearest tenth)
l & h are the lower and higher footprint limits or constraints or
(``kinks'') or the boundary between the flat regions and the
intermediate sloped line (in sq ft)

Graphically, piecewise linear form, like the constrained linear
form, appears as shown in Figure II.C-2.

[[Page 25358]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.006

BILLING CODE 6560-50-C

[[Page 25359]]

As for the constrained linear form, the specific form and
stringency of the piecewise linear function for each fleet (passenger
car and light trucks) and model year are defined through specific
values for the four coefficients shown above.
For purposes of the proposed rules, NHTSA and EPA developed the
basic curve shapes using methods similar to those applied by NHTSA in
fitting the curves defining the MY 2011 standards. The first step
involved defining the relevant vehicle characteristics in the form used
by NHTSA's CAFE model (e.g., fuel economy, footprint, vehicle class,
technology) described in Section II.B of this preamble and in Chapter 1
of the Joint TSD. However, because the baseline fleet utilizes a wide
range of available fuel saving technologies, NHTSA used the CAFE model
to develop a fleet to which all of the technologies discussed in
Chapter 3 of the Joint TSD \63\ were applied, except dieselization and
strong hybridization. This was accomplished by taking the following
steps: (1) Treating all manufacturers as unwilling to pay civil
penalties rather than applying technology, (2) applying any technology
at any time, irrespective of scheduled vehicle redesigns or freshening,
and (3) ignoring ``phase-in caps'' that constrain the overall amount of
technology that can be applied by the model to a given manufacturer's
fleet. These steps helped to increase technological parity among
vehicle models, thereby providing a better basis (than the baseline or
reference fleets) for estimating the statistical relationship between
vehicle size and fuel economy.
---------------------------------------------------------------------------

\63\ The agencies excluded diesel engines and strong hybrid
vehicle technologies from this exercise (and only this exercise)
because the agencies expect that manufacturers would not need to
rely heavily on these technologies in order to comply with the
proposed standards. NHTSA and EPA did include diesel engines and
strong hybrid vehicle technologies in all other portions of their
analyses.
---------------------------------------------------------------------------

In fitting the curves, NHTSA and EPA also continued to fit the
sloped portion of the function to vehicle models between the footprint
values at which the agencies continued to apply constraints to limit
the function's value for both the smallest and largest vehicles.
Without a limit at the smallest footprints, the function--whether
logistic or linear--can reach values that would be unfairly burdensome
for a manufacturer that elects to focus on the market for small
vehicles; depending on the underlying data, an unconstrained form,
could result in stringency levels that are technologically infeasible
and/or economically impracticable for those manufacturers that may
elect to focus on the smallest vehicles. On the other side of the
function, without a limit at the largest footprints, the function may
provide no floor on required fuel economy. Also, the safety
considerations that support the provision of a disincentive for
downsizing as a compliance strategy apply weakly, if at all, to the
very largest vehicles. Limiting the function's value for the largest
vehicles leads to a function with an inherent absolute minimum level of
performance, while remaining consistent with safety considerations.
Before fitting the sloped portion of the constrained linear form,
NHTSA and EPA selected footprints above and below which to apply
constraints (i.e., minimum and maximum values) on the function. The
agencies believe that the linear form performs well in describing the
observed relationship between footprint and fuel consumption or
CO2 emissions for vehicle models within the footprint ranges
covering most vehicle models, but that the single (as opposed to
piecewise) linear form does not perform well in describing this
relationship for the smallest and largest vehicle models. For passenger
cars, the agency noted that several manufacturers offer small, sporty
coupes below 41 square feet, such as the BMW Z4 and Mini, Honda S2000,
Mazda MX-5 Miata, Porsche Carrera and 911, and Volkswagen New Beetle.
Because such vehicles represent a small portion (less than 10 percent)
of the passenger car market, yet often have performance, utility, and/
or structural characteristics that could make it technologically
infeasible and/or economically impracticable for manufacturers focusing
on such vehicles to achieve the very challenging average requirements
that could apply in the absence of a constraint, EPA and NHTSA proposed
to ``cut off'' the linear portion of the passenger car function at 41
square feet. The agencies recognize that for manufacturers who make
small vehicles in this size range, this cut off creates some incentive
to downsize (i.e., further reduce the size, and/or increase the
production of models currently smaller than 41 square feet) to make it
easier to meet the target. The cut off may also create the incentive
for manufacturers who do not currently offer such models to do so in
the future. However, at the same time, the agencies believe that there
is a limit to the market for cars smaller than 41 square feet--most
consumers likely have some minimum expectation about interior volume,
among other things. The agencies thus believe that the number of
consumers who will want vehicles smaller than 41 square feet
(regardless of how they are priced) is small, and that the incentive to
downsize in response to this final rule, if present, will be minimal.
For consistency, the agency proposed to ``cut off'' the light truck
function at the same footprint, although no light trucks are currently
offered below 41 square feet. The agencies further noted that above 56
square feet, the only passenger car model present in the MY 2008 fleet
were four luxury vehicles with extremely low sales volumes--the Bentley
Arnage and three versions of the Rolls Royce Phantom. NHTSA and EPA
therefore also proposed to ``cut off'' the linear portion of the
passenger car function at 56 square feet. Finally, the agencies noted
that although public information is limited regarding the sales volumes
of the many different configurations (cab designs and bed sizes) of
pickup trucks, most of the largest pickups (e.g., the Ford F-150, GM
Sierra/Silverado, Nissan Titan, and Toyota Tundra) appear to fall just
above 66 square feet in footprint. EPA and NHTSA therefore proposed to
``cut off'' the linear portion of the light truck function at 66 square
feet.
Having developed a set of vehicle emissions and footprint data
which represent the benefit of all non-diesel, non-hybrid technologies,
we determined the initial values for parameters c and d were determined
for cars and trucks separately. c and d were initially set at the
values for which the average (equivalently, sum) of the absolute values
of the differences was minimized between the ``maximum technology''
fleet fuel consumption (within the footprints between the upper and
lower limits) and the straight line of the function defined above at
the same corresponding vehicle footprints. That is, c and d were
determined by minimizing the average absolute residual, commonly known
as the MAD (Mean Absolute Deviation) approach, of the corresponding
straight line.
Finally, NHTSA calculated the values of the upper and lower
parameters (a and b) based on the corresponding footprints discussed
above (41 and 56 square feet for passenger cars, and 41 and 66 square
feet for light trucks).
The result of this methodology is shown below in Figures II.C-3 and
II.C-4 for passenger cars and light trucks, respectively. The fitted
curves are shown with the underlying ``maximum technology'' passenger
car and light truck fleets. For passenger cars, the mean absolute
deviation of the sloped portion of the function was 14 percent.

[[Page 25360]]

For trucks, the corresponding MAD was 10 percent.
BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TR07MY10.007

[[Page 25361]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.008

BILLING CODE 6560-50-C

[[Page 25362]]

The agencies used these functional forms as a starting point to
develop mathematical functions defining the actual proposed standards
as discussed above. The agencies then transposed these functions
vertically (i.e., on a gpm or CO2 basis, uniformly downward)
to produce the same fleetwide fuel economy (and CO2 emission
levels) for cars and light trucks described in the NPRM.
A number of public comments generally supported the agencies'
choice of attribute-based mathematical functions, as well as the
methods applied to fit the function. Ferrari indicated support for the
use of a constrained linear form rather than a constrained logistic
form, support for the application of limits on the functions' values,
support for a generally less steep passenger car curve compared to MY
2011, and support for the inclusion of all manufacturers in the
analysis used to fit the curves. ICCT also supported the use of a
constrained linear form. Toyota expressed general support for the
methods and outcome, including a less-steep passenger car curve, and
the application of limits on fuel economy targets applicable to the
smallest vehicles. The UAW commented that the shapes and levels of the
curves are reasonable.
Other commenters suggested that changes to the agencies' methods
and results would yield better outcomes. GM suggested that steeper
curves would provide a greater incentive for limited-line manufacturers
to apply technology to smaller vehicles. GM argued that steeper and, in
their view, fairer curves could be obtained by using sales-weighted
least-squares regression rather than minimization of the unweighted
mean absolute deviation. Conversely, students from UC Santa Barbara
commented that the passenger car and light truck curves should be
flatter and should converge over time in order to encourage the market
to turn, as the agencies' analysis assumes it will, away from light
trucks and toward passenger cars.
NADA commented that there should be no ``cut-off'' points (i.e.,
lower limits or floors), because these de facto ``backstops'' might
limit consumer choice, especially for light trucks--a possibility also
suggested by the Alliance. The Alliance and several individual
manufacturers also commented that the cut-off point for light trucks
should be shifted to 72 square feet (from the proposed 66 square feet),
arguing that the preponderance of high-volume light truck models with
footprints greater than 66 square feet is such that a 72 square foot
cut-off point makes it unduly challenging for manufacturers serving the
large pickup market and thereby constitutes a de facto backstop. Also,
with respect to the smallest light truck models, Honda commented that
the cut-off point should be set at the point defining the smallest 10
percent of the fleet, both for consistency with the passenger car cut-
off point, and to provide a greater incentive for manufacturers to
downsize the smallest light truck models (which provide greater
functionality than passenger cars).
Other commenters focused on whether the agencies should have
separate curves for different fleets or whether they should have a
single curve that applied to both passenger cars and light trucks. This
issue is related, to some extent, to commenters who discussed whether
car and truck definitions should change. CARB, Ford, and Toyota
supported separate curves for cars and trucks, generally stating that
different fleets have different functional characteristics and these
characteristics are appropriately addressed by separate curves.
Likewise, AIAM, Chrysler, and NADA supported leaving the current
definitions of car and truck the same. CBD, ICCT, and NESCAUM supported
a single curve, based on concerns about manufacturers gaming the system
and reclassifying passenger cars as light trucks in order to obtain the
often-less stringent light truck standard, which could lead to lower
benefits than anticipated by the agencies.
In addition, the students from UC Santa Barbara reported being
unable to reproduce the agencies' analysis to fit curves to the
passenger car and light truck fleets, even when using the model,
inputs, and external analysis files posted to NHTSA's Web site when the
NPRM was issued.
Having considered public comments, NHTSA and EPA have re-examined
the development of curves underlying the standards proposed in the
NPRM, and are promulgating standards based on the same underlying
curves. The agencies have made this decision considering that, while
EISA mandates that CAFE standards be defined by a mathematical function
in terms of one or more attributes related to fuel economy, neither
EISA nor the CAA require that the mathematical function be limited to
the observed or theoretical dependence of fuel economy on the selected
attribute or attributes. As a means by which CAFE and GHG standards are
specified, the mathematical function can and does properly play a
normative role. Therefore, NHTSA and EPA have concluded that, as
supported by comments, the mathematical function can reasonably be
based on a blend of analytical and policy considerations, as discussed
below and in the Joint Technical Support Document.
With respect to GM's recommendation that NHTSA and EPA use weighted
least-squares analysis, the agencies find that the market forecast used
for analysis supporting both the NPRM and the final rule exhibits the
two key characteristics that previously led NHTSA to use minimization
of the unweighted Mean Absolute Deviation (MAD) rather than weighted
least-squares analysis. First, projected model-specific sales volumes
in the agencies' market forecast cover an extremely wide range, such
that, as discussed in NHTSA's rulemaking for MY 2011, while unweighted
regression gives low-selling vehicle models and high-selling vehicle
models equal emphasis, sales-weighted regression would give some
vehicle models considerably more emphasis than other vehicle
models.\64\ The agencies' intention is to fit a curve that describes a
technical relationship between fuel economy and footprint, given
comparable levels of technology, and this supports weighting discrete
vehicle models equally. On the other hand, sales weighted regression
would allow the difference between other vehicle attributes to be
reflected in the analysis, and also would reflect consumer demand.
---------------------------------------------------------------------------

\64\ For example, the agencies' market forecast shows MY 2016
sales of 187,000 units for Toyota's 2WD Sienna, and shows 27 model
configurations with MY 2016 sales of fewer than 100 units.
Similarly, the agencies' market forecast shows MY 2016 sales of
268,000 for the Toyota Prius, and shows 29 model configurations with
MY 2016 sales of fewer than 100 units. Sales-weighted analysis would
give the Toyota Sienna and Prius more than a thousand times the
consideration of many vehicle model configurations. Sales-weighted
analysis would, therefore, cause a large number of vehicle model
configurations to be virtually ignored. See discussion in NHTSA's
final rule for MY 2011 passenger car and light truck CAFE standards,
74 FR 14368 (Mar. 30, 2009), and in NHTSA's NPRM for that
rulemaking, 73 FR 24423-24429 (May 2, 2008).
---------------------------------------------------------------------------

Second, even after NHTSA's ``maximum technology'' analysis to
increase technological parity of vehicle models before fitting curves,
the agencies' market forecast contains many significant outliers. As
discussed in NHTSA's rulemaking for MY 2011, MAD is a statistical
procedure that has been demonstrated to produce more efficient
parameter estimates than least-squares analysis in the presence of
significant outliers.\65\ In addition, the

[[Page 25363]]

agencies remain concerned that the steeper curves resulting from
weighted least-squares analysis would increase the risk that energy
savings and environmental benefits would be lower than projected,
because the steeper curves would provide a greater incentive to
increase sales of larger vehicles with lower fuel economy levels. Based
on these technical considerations and these concerns regarding
potential outcomes, the agencies have decided not to re-fit curves
using weighted least-squares analysis, but note that they may
reconsider using least-squares regression in future analysis.
---------------------------------------------------------------------------

\65\ Id. In the case of a dataset not drawn from a sample with a
Gaussian, or normal, distribution, there is often a need to employ
robust estimation methods rather than rely on least-squares approach
to curve fitting. The least-squares approach has as an underlying
assumption that the data are drawn from a normal distribution, and
hence fits a curve using a sum-of-squares method to minimize errors.
This approach will, in a sample drawn from a non-normal
distribution, give excessive weight to outliers by making their
presence felt in proportion to the square of their distance from the
fitted curve, and, hence, distort the resulting fit. With outliers
in the sample, the typical solution is to use a robust method such
as a minimum absolute deviation, rather than a squared term, to
estimate the fit (see, e.g., ``AI Access: Your Access to Data
Modeling,'' at http://www.aiaccess.net/English/Glossaries/GlosMod/
e_gm_O_Pa.htm#Outlier). The effect on the estimation is to let
the presence of each observation be felt more uniformly, resulting
in a curve more representative of the data (see, e.g., Peter
Kennedy, A Guide to Econometrics, 3rd edition, 1992, MIT Press,
Cambridge, MA).
---------------------------------------------------------------------------

NHTSA and EPA have considered GM's comment that steeper curves
would provide a greater incentive for limited-line manufacturers to
apply technology to smaller vehicles. While the agencies agree that a
steeper curve would, absent any changes in fleet mix, tend to shift
average compliance burdens away from GM and toward companies that make
smaller vehicles, the agencies are concerned, as stated above, that
steeper curves would increase the risk that induced increases in
vehicle size could erode projected energy and environmental benefits.
NHTSA and EPA have also considered the comments by the students
from UC Santa Barbara indicating that the passenger car and light truck
curves should be flatter and should converge over time. The agencies
conclude that flatter curves would reduce the incentives intended in
shifting from ``flat'' CAFE standards to attribute-based CAFE and GHG
standards--those being the incentive to respond to attribute-based
standards in ways that minimize compromises in vehicle safety, and the
incentive for more manufacturers (than primarily those selling a wider
range of vehicles) across the range of the attribute to have to
increase the application of fuel-saving technologies. With regard to
whether the agencies should set separate curves or a single one, NHTSA
also notes that EPCA requires NHTSA to establish standards separately
for passenger cars and light trucks, and thus concludes that the
standards for each fleet should be based on the characteristics of
vehicles in each fleet. In other words, the passenger car curve should
be based on the characteristics of passenger cars, and the light truck
curve should be based on the characteristics of light trucks--thus to
the extent that those characteristics are different, an artificially-
forced convergence would not accurately reflect those differences.
However, such convergence could be appropriate depending on future
trends in the light vehicle market, specifically further reduction in
the differences between passenger car and light truck characteristics.
While that trend was more apparent when car-like 2WD SUVs were
classified as light trucks, it seems likely to diminish for the model
year vehicles subject to these rules as the truck fleet will be more
purely ``truck-like'' than has been the case in recent years.
NHTSA and EPA have also considered comments on the maxima and
minima that the agencies have applied to ``cut off'' the linear
function underlying the proposed curves for passenger cars and light
trucks. Contrary to NADA's suggestion that there should be no such cut-
off points, the agencies conclude that curves lacking maximum fuel
economy targets (i.e., minimum CO2 targets) would result in
average fuel economy and GHG requirements that would not be
technologically feasible or economically practicable for manufacturers
concentrating on those market segments. In addition, minimum fuel
economy targets (i.e., maximum CO2 targets) are important to
mitigate the risk to energy and environmental benefits of potential
market shifts toward large vehicles. The agencies also disagree with
comments by the Alliance and several individual manufacturers that the
cut-off point for light trucks should be shifted to 72 square feet
(from the proposed 66 square feet) to ease compliance burdens facing
manufacturers serving the large pickup market. Such a shift would
increase the risk that energy and environmental benefits of the
standards would be compromised by induced increases in the sales of
large pickups, in situations where the increased compliance burden is
feasible and appropriate. Also, the agencies' market forecast suggests
that most of the light trucks models with footprints larger than 66
square feet have curb weights near or above 5,000 pounds. This
suggests, in turn, that in terms of highway safety, there is little or
no need to discourage downsizing of light trucks with footprints larger
than 66 square feet. Based on these energy, environmental,
technological feasibility, economic practicability, and safety
considerations, the agencies conclude that the light truck curve should
be cut off at 66 square feet, as proposed, rather than at 72 square
feet. The agencies also disagree with Honda's suggestion that the cut-
off point for the smallest trucks be shifted to a larger footprint
value, because doing so could potentially increase the incentive to
reclassify vehicles in that size range as light trucks, and could
thereby increase the possibility that energy and environmental benefits
of the rule would be less than projected.
Finally, considering comments by the UC Santa Barbara students
regarding difficulties reproducing NHTSA's analysis, NHTSA reexamined
its analysis, and discovered some erroneous entries in model inputs
underlying the analysis used to develop the curves proposed in the
NPRM. These errors are discussed in NHTSA's final Regulatory Impact
Analysis (FRIA) and have since been corrected. They include the
following: Incorrect valvetrain phasing and lift inputs for many BMW
engines, incorrect indexing for some Daimler models, incorrectly
enabled valvetrain technologies for rotary engines and Atkinson cycle
engines, omitted baseline applications of cylinder deactivation in some
Honda and GM engines, incorrect valve phasing codes for some 4-cylinder
Chrysler engines, omitted baseline applications of advanced
transmissions in some VW models, incorrectly enabled advanced
electrification technologies for several hybrid vehicle models, and
incorrect DCT effectiveness estimates for subcompact passenger cars.
These errors, while not significant enough to impact the overall
analysis of stringency, did affect the fitted slope for the passenger
car curve and would have prevented precise replication of NHTSA's NPRM
analysis by outside parties.
After correcting these errors and repeating the curve development
analysis presented in the NPRM, NHTSA obtained the curves shown below
in Figures II.C-5 and II.C-6 for passenger cars and light trucks,
respectively. The fitted curves are shown with the underlying ``maximum
technology'' passenger car and light truck fleets. For passenger cars,
the mean absolute deviation of the sloped portion of the function was
14 percent. For trucks, the corresponding MAD was 10 percent.
BILLING CODE 6560-50-P

[[Page 25364]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.009

[[Page 25365]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.010

[[Page 25366]]

This refitted passenger car curve is similar to that presented in
the NPRM, and the refitted light truck curve is nearly identical to the
corresponding curve in the NPRM. However, the slope of the refitted
passenger car curve is about 27 percent steeper (on a gpm per sf basis)
than the curve presented in the NPRM. For passenger cars and light
trucks, respectively, Figures II.C-7 and II.C-8 show the results of
adjustment--discussed in the next section--of the above curves to yield
the average required fuel economy levels corresponding to the final
standards.
[GRAPHIC] [TIFF OMITTED] TR07MY10.011

[[Page 25367]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.012

BILLING CODE 6560-50-C
While the resultant light truck curves are visually
indistinguishable from one another, the refitted curve for passenger
cars would increase stringency for the smallest cars, decrease
stringency for the largest cars, and provide a greater incentive to
increase vehicle size throughout the range of footprints within which
NHTSA and EPA project most passenger car models will be sold through MY
2016. The agencies are concerned that these changes would make it
unduly difficult for manufacturers to introduce new small passenger
cars in the United States, and unduly risk losses in energy and
environmental benefits by increasing incentives for the passenger car
market to shift toward larger vehicles.
Also, the agencies note that the refitted passenger car curve
produces only a slightly closer fit to the corrected fleet than would
the curve estimated in

[[Page 25368]]

the NPRM; with respect to the corrected fleet (between the ``cut off''
footprint values, and after the ``maximum technology'' analysis
discussed above), the mean absolute deviation for the refitted curve is
13.887 percent, and that of a refitted curve held to the original slope
is 13.933 percent. In other words, the data support the original slope
very nearly as well as they support the refitted slope.
Considering NHTSA's and EPA's concerns regarding the change in
incentives that would result from a refitted curve for passenger cars,
and considering that the data support the original curves about as well
as they would support refitted curves, the agencies are finalizing CAFE
and GHG standards based on the curves presented in the NPRM.
Finally, regarding some commenters' inability to reproduce the
agencies' NPRM analysis, NHTSA believes that its correction of the
errors discussed above and its release (on NHTSA's Web site) of the
updated Volpe model and all accompanying inputs and external analysis
files should enable outside parties to independently reproduce the
agencies' analysis. If outside parties continue to experience
difficulty in doing so, we encourage them to contact NHTSA, and the
agency will do its best to provide assistance.
Thus, in summary, the agencies' approach to developing the
attribute-based mathematical functions for MY 2012-2016 CAFE and
CO2 standards represents the agencies' best technical
judgment and consideration of potential outcomes at this time, and we
are confident that the conclusions have resulted in appropriate and
reasonable standards. The agencies recognize, however, that aspects of
these decisions may merit updating or revision in future analysis to
support CAFE and CO2 standards or for other purposes.
Consistent with best rulemaking practices, the agencies will take a
fresh look at all assumptions and approaches to curve fitting,
appropriate attributes, and mathematical functions in the context of
future rulemakings.
The agencies also recognized in the NPRM the possibility that lower
fuel prices could lead to lower fleetwide fuel economy (and higher
CO2 emissions) than projected in this rule. One way of
addressing that concern is through the use of a universal standard--
that is, an average standard set at a (single) absolute level. This is
often described as a ``backstop standard.'' The agencies explained that
under the CAFE program, EISA requires such a minimum average fuel
economy standard for domestic passenger cars, but is silent with regard
to similar backstops for imported passenger cars and light trucks,
while under the CAA, a backstop could be adopted under section 202(a)
assuming it could be justified under the relevant statutory criteria.
NHTSA and EPA also noted that the flattened portions of the curves at
the largest footprints directionally address the issue of a backstop
(i.e., the mpg ``floor'' or gpm ``ceiling'' applied to the curves
provides a universal and absolute value for that range of footprints).
The agencies sought comment on whether backstop standards, or any other
method within the agencies' statutory authority, should and can be
implemented in order to guarantee a level of CO2 emissions
reductions and fuel savings under the attribute-based standards.
The agencies received a number of comments regarding the need for a
backstop beyond NHTSA's alternative minimum standard. Comments were
divided fairly evenly between support for and opposition to additional
backstop standards. The following organizations supported the need for
EPA and NHTSA to have explicit backstop standards: American Council for
an Energy Efficient Economy (ACEEE), American Lung Association,
California Air Resources Board (CARB), Environment America, Environment
Defense Fund, Massachusetts Department of Environmental Protection,
Natural Resources Defense Council (NRDC), Northeast States for
Coordinated Air Use Management (NESCAUM), Public Citizen and Safe
Climate Campaign, Sierra Club, State of Washington Department of
Ecology, Union of Concerned Scientists, and a number of private
citizens. Commenters in favor of additional backstop standards for all
fleets for both NHTSA and EPA \66\ generally stated that the emissions
reductions and fuel savings expected to be achieved by MY 2016 depended
on assumptions about fleet mix that might not come to pass, and that
various kinds of backstop standards or ``ratchet mechanisms'' \67\ were
necessary to ensure that those reductions were achieved in fact. In
addition, some commenters \68\ stated that manufacturers might build
larger vehicles or more trucks during MYs 2012-2016 than the agencies
project, for example, because (1) any amount of slope in target curves
encourages manufacturers to upsize, and (2) lower targets for light
trucks than for passenger cars encourage manufacturers to find ways to
reclassify vehicles as light trucks, such as by dropping 2WD versions
of SUVs and offering only 4WD versions, perhaps spurred by NHTSA's
reclassification of 2WD SUVs as passenger cars. Both of these
mechanisms will be addressed further below. Some commenters also
discussed EPA authority under the CAA to set backstops,\69\ agreeing
with EPA's analysis that section 202(a) allows such standards since EPA
has wide discretion under that section to craft standards.
---------------------------------------------------------------------------

\66\ ACEEE, American Lung Association, CARB, Christopher Lish,
Environment America, EDF, MA DEP, NRDC, NESCAUM, Public Citizen,
Sierra Club et al., SCAQMD, UCS, WA DE.
\67\ Commenters generally defined a ``ratchet mechanism'' as an
automatic re-calculation of stringency to ensure cumulative goals
are reached by 2016, even if emissions reductions and fuel savings
fall short in the earlier years covered by the rulemaking.
\68\ CBD, MA DEP, NJ DEP, Public Citizen, Sierra Club et al.,
UCS.
\69\ CARB, Public Citizen, Sierra Club et al.
---------------------------------------------------------------------------

The following organizations opposed a backstop: Alliance of
Automobile Manufacturers (AAM), Association of International Automobile
Manufacturers (AIAM), Ford Motor Company, National Automobile Dealers
Association (NADA), Toyota Motor Company, and the United Auto Workers
Union. Commenters stating that additional backstops would not be
necessary disagreed that upsizing was likely,\70\ and emphasized the
anti-backsliding characteristics of the target curves. Others argued
that universal absolute standards as backstops could restrict consumer
choice of vehicles. Commenters making legal arguments under EPCA/
EISA\71\ stated that Congress' silence regarding backstops for imported
passenger cars and light trucks should be construed as a lack of
authority for NHTSA to create further backstops. Commenters making
legal arguments under the CAA\72\ focused on the lack of clear
authority under the CAA to create multiple GHG emissions standards for
the same fleets of vehicles based on the same statutory criteria, and
opposed EPA taking steps that would reduce harmonization with NHTSA in
standard setting. Furthermore, AIAM indicated that EISA's requirement
that the combined (car and truck) fuel economy level reach at least 35
mpg by

[[Page 25369]]

2020 itself constitutes a backstop.\73\ One individual \74\ commented
that while additional backstop standards might be necessary given
optimism of fleet mix assumptions, both agencies' authorities would
probably need to be revised by Congress to clarify that backstop
standards (whether for individual fleets or for the national fleet as a
whole) were permissible.
---------------------------------------------------------------------------

\70\ For example, the Alliance and Toyota said that upsizing
would not be likely because (1) it would not necessarily make
compliance with applicable standards easier, since larger vehicles
tend to be heavier and heavier vehicles tend to achieve worse fuel
economy/emissions levels; (2) it may require expensive platform
changes; (3) target curves become increasingly more stringent from
year to year, which reduces the benefits of upsizing; and (4) the
mpg floor and gpm ceiling for the largest vehicles (the point at
which the curve is ``cut off'') discourages manufacturers from
continuing to upsize beyond a point because doing so makes it
increasingly difficult to meet the flat standard at that part of the
curve.
\71\ AIAM, Alliance, Ford, NADA, Toyota.
\72\ Alliance, Ford, NADA, UAW.
\73\ NHTSA and EPA agree with AIAM that the EISA 35 mpg
requirement in MY 2020 has a backstop-like function, in that it
requires a certain level of achieved fleetwide fuel economy by a
certain date, although it is not literally a backstop standard.
Considering that NHTSA's MY 2011 CAFE standards increased projected
average fuel economy requirements (relative to the MY 2010
standards) at a significantly faster rate than would be required to
achieve the 35-in-2020 requirement, and considering that the
standards being finalized today would increase projected average
combined fuel economy requirements to 34.1 mpg in MY 2016, four
years before MY 2020, the agencies believe that the U.S. vehicle
market would have to shift in highly unexpected ways in order to put
the 35-in-2020 requirement at risk, even despite the fact that due
to the attribute-based standards, average fuel economy requirements
will vary depending on the mix of vehicles produced for sale in the
U.S. in each model year. The agencies further emphasize that both
NHTSA and EPA plan to conduct and document retrospective analyses to
evaluate how the market's evolution during the rulemaking timeframe
compares with the agencies' forecasts employed for this rulemaking.
Additionally, we emphasize that both agencies have the authority,
given sufficient lead time, to revise their standards upwards if
necessary to avoid missing the 35-in-2020 requirement.
\74\ Schade.
---------------------------------------------------------------------------

In response, EPA and NHTSA remain confident that their projections
of the future fleet mix are reliable, and that future changes in the
fleet mix of footprints and sales are not likely to lead to more than
modest changes in projected emissions reductions or fuel savings.\75\
Both agencies thus remain confident in these fleet projections and the
resulting emissions reductions and fuel savings from the standards. As
explained in Section II.B above, the agencies' projections of the
future fleet are based on the most transparent information currently
available to the agencies. In addition, there are only a relatively few
model years at issue. Moreover, market trends today are consistent with
the agencies' estimates, showing shifts from light trucks to passenger
cars and increased emphasis on fuel economy from all vehicles.
---------------------------------------------------------------------------

\75\ For reference, NHTSA's March 2009 final rule establishing
MY 2011 CAFE standards was based on a forecast that passenger cars
would represent 57.6 percent of the MY 2011 fleet, and that MY 2011
passenger cars and light trucks would average 45.6 square feet (sf)
and 55.1 sf, respectively, such that average required CAFE levels
would be 30.2 mpg, 24.1 mpg, and 27.3 mpg, respectively, for
passenger cars, light trucks, and the overall light-duty fleet.
Based on the agencies' current market forecast, even as soon as MY
2011, passenger cars will comprise a larger share (59.2 percent) of
the light vehicle market; passenger cars and light trucks will, on
average, be smaller by 0.5 sf and 1.3 sf, respectively; and average
required CAFE levels will be higher by 0.2 mpg, 0.3 mpg, and 0.3
mpg, respectively, for passenger cars, light trucks, and the overall
light-duty fleet.
---------------------------------------------------------------------------

Finally, the shapes of the curves, including the ``flattening'' at
the largest footprint values, tend to avoid or minimize regulatory
incentives for manufacturers to upsize their fleet to change their
compliance burden. Given the way the curves are fit to the data points
(which represent vehicle models' fuel economy mapped against their
footprint), the agencies believe that there is little real benefit to
be gained by a manufacturer upsizing their vehicles. As discussed
above, the agencies' analysis indicates that, for passenger car models
with footprints falling between the two flattened portions of the
corresponding curve, the actual slope of fuel economy with respect to
footprint, if fit to that data by itself, is about 27 percent steeper
than the curve the agencies are promulgating today. This difference
suggests that manufacturers would, if anything, have more to gain by
reducing vehicle footprint than by increasing vehicle footprint. For
light trucks, the agencies' analysis indicates that, for models with
footprints falling between the two flatted portions of the
corresponding curve, the slope of fuel economy with respect to
footprint is nearly identical to the curve the agencies are
promulgating today. This suggests that, within this range,
manufacturers would typically have little incentive to either
incrementally increase or reduce vehicle footprint. The agencies
recognize that based on economic and consumer demand factors that are
external to this rule, the distribution of footprints in the future may
be different (either smaller or larger) than what is projected in this
rule. However, the agencies continue to believe that there will not be
significant shifts in this distribution as a direct consequence of this
rule.
At the same time, adding another backstop standard would have
virtually no effect if the standard was weak, but a more stringent
backstop could compromise the objectives served by attribute-based
standards--that they distribute compliance burdens more equally among
manufacturers, and at the same time encourage manufacturers to apply
fuel-saving technologies rather than simply downsizing their vehicles,
as they did in past decades under flat standards. This is why Congress
mandated attribute-based CAFE standards in EISA. This compromise in
objectives could occur for any manufacturer whose fleet average was
above the backstop, irrespective of why they were above the backstop
and irrespective of whether the industry as a whole was achieving the
emissions and fuel economy benefits projected for the final standards,
the problem the backstop is supposed to address. For example, the
projected industry wide level of 250 gm/mile for MY 2016 is based on a
mix of manufacturer levels, ranging from approximately 205 to 315 gram/
mile \76\ but resulting in an industry wide basis in a fleet average of
250 gm/mile. Unless the backstop was at a very weak level, above the
high end of this range, then some percentage of manufacturers would be
above the backstop even if the performance of the entire industry
remains fully consistent with the emissions and fuel economy levels
projected for the final standards. For these manufacturers and any
other manufacturers who were above the backstop, the objectives of an
attribute based standard would be compromised and unnecessary costs
would be imposed. This could directionally impose increased costs for
some manufacturers. It would be difficult if not impossible to
establish the level of a backstop standard such that costs are likely
to be imposed on manufacturers only when there is a failure to achieve
the projected reductions across the industry as a whole. An example of
this kind of industry wide situation could be when there is a
significant shift to larger vehicles across the industry as a whole, or
if there is a general market shift from cars to trucks. The problem the
agencies are concerned about in those circumstances is not with respect
to any single manufacturer, but rather is based on concerns over shifts
across the fleet as a whole, as compared to shifts in one
manufacturer's fleet that may be more than offset by shifts the other
way in another manufacturer's fleet. However, in this respect, a
traditional backstop acts as a manufacturer specific standard.
---------------------------------------------------------------------------

\76\ Based on estimated standards presented in Tables III.B.1-1
and III.B.1-2.
---------------------------------------------------------------------------

The concept of a ratchet mechanism recognizes this problem, and
would impose the new more stringent standard only when the problem
arises across the industry as a whole. While the new more stringent
standards would enter into force automatically, any such standards
would still need to provide adequate lead time for the manufacturers.
Given the limited number of model years covered by this rulemaking and
the short lead-time already before the 2012 model year, a ratchet
mechanism in this rulemaking that would automatically tighten the
standards at some point after model year 2012 is finished and apply the
new more stringent standards for model

[[Page 25370]]

years 2016 or earlier, would fail to provide adequate lead time for any
new, more stringent standards
Additionally, we do not believe that the risk of vehicle upsizing
or changing vehicle offerings to ``game'' the passenger car and light
truck definitions is as great as commenters imply for the model years
in question.\77\ The changes that commenters suggest manufacturers
might make are neither so simple nor so likely to be accepted by
consumers. For example, 4WD versions of vehicles tend to be more
expensive and, other things being equal, have inherently lower fuel
economy than their 2WD equivalent models. Therefore, although there is
a market for 4WD vehicles, and some consumers might shift from 2WD
vehicles to 4WD vehicles if 4WD becomes available at little or no extra
cost, many consumers still may not desire to purchase 4WD vehicles
because of concerns about cost premium and additional maintenance
requirements; conversely, many manufacturers often require the 2WD
option to satisfy demand for base vehicle models. Additionally,
increasing the footprint of vehicles requires platform changes, which
usually requires a product redesign phase (the agencies estimate that
this occurs on average once every 5 years for most models).
Alternatively, turning many 2WD SUVs into 2WD light trucks would
require manufacturers to squeeze a third row of seats in or
significantly increase their GVWR, which also requires a significant
change in the vehicle.\78\ The agencies are confident that the
anticipated increases in average fuel economy and reductions in average
CO2 emission rates can be achieved without backstops under
EISA or the CAA. As noted above, the agencies plan to conduct
retrospective analysis to monitor progress. Both agencies have the
authority to revise standards if warranted, as long as sufficient lead
time is provided.
---------------------------------------------------------------------------

\77\ We note that NHTSA's recent clarification of the light
truck definitions has significantly reduced the potential for
gaming, and resulted in the reclassification of over a million
vehicles from the light truck to the passenger car fleet.
\78\ Increasing the GVWR of a light truck (assuming this was the
only goal) can be accomplished in a number of ways, and must include
consideration of: (1) Redesign of wheel axles; (2) improving the
vehicle suspension; (3) changes in tire specification (which will
likely affect ride quality); (4) vehicle dynamics development
(especially with vehicles equipped with electronic stability
control); and (5) brake redesign. Depending on the vehicle, some of
these changes may be easier or more difficult than others.
---------------------------------------------------------------------------

The agencies acknowledge that the MY 2016 fleet emissions and fuel
economy goals of 250 g/mi and 34.1 mpg for EPA and NHTSA respectively
are estimates and not standards (the MY 2012-2016 curves are the
standards). Changes in fuel prices, consumer preferences, and/or
vehicle survival and mileage accumulation rates could result in either
smaller or larger oil and GHG savings. As explained above and elsewhere
in the rule, the agencies believe that the possibility of not meeting
(or, alternatively, exceeding) fuel economy and emissions goals exists,
but is not likely. Given this, and given the potential complexities in
designing an appropriate backstop, the agencies believe the balance
here points to not adopting additional backstops at this time for the
MYs 2012-2016 standards other than NHTSA's finalizing of the ones
required by EPCA/EISA for domestic passenger cars. Nevertheless, the
agencies recognize there are many factors that are inherently uncertain
which can affect projections in the future, including fuel price and
other factors which are unrelated to the standards contained in this
final rule. Such factors can affect consumer preferences and are
difficult to predict. At this time and based on the available
information, the agencies have not included a backstop for model years
2012-2016. However, if circumstances change in the future in
unanticipated ways, the agencies may revisit the issue of a backstop in
the context of a future rulemaking either for model years 2012-2016 or
as needed for standards for model years beyond 2016. This issue will be
discussed further in Sections III and IV.

D. Relative Car-Truck Stringency

The agencies proposed fleetwide standards with the projected levels
of stringency of 34.1 mpg or 250 g/mi in MY 2016 (as well as the
corresponding intermediate year fleetwide standards) for NHTSA and EPA
respectively. To determine the relative stringency of passenger car and
light truck standards for those model years, the agencies were
concerned that increasing the difference between the car and truck
standards (either by raising the car standards or lowering the truck
standards) could encourage manufacturers to build fewer cars and more
trucks, likely to the detriment of fuel economy and CO2
reductions.\79\ In order to maintain consistent car/truck standards,
the agencies applied a constant ratio between the estimated average
required performance under the passenger car and light truck standards,
in order to maintain a stable set of incentives regarding vehicle
classification.
---------------------------------------------------------------------------

\79\ For example, since many 2WD SUVs are classified as
passenger cars, manufacturers have already warned that high car
standards relative to truck standards could create an incentive for
them to drop the 2WD version and sell only the 4WD version.
---------------------------------------------------------------------------

To calculate relative car-truck stringency for the proposal, the
agencies explored a number of possible alternatives, and for the
reasons described in the proposal used the Volpe model in order to
estimate stringencies at which net benefits would be maximized. The
agencies have followed the same approach in calculating the relative
car-truck stringency for the final standards promulgated today. Further
details of the development of this approach can be found in Section IV
of this preamble as well as in NHTSA's RIA and EIS. NHTSA examined
passenger car and light truck standards that would produce the proposed
combined average fuel economy levels from Table I.B.2-2 above. NHTSA
did so by shifting downward the curves that maximize net benefits,
holding the relative stringency of passenger car and light truck
standards constant at the level determined by maximizing net benefits,
such that the average fuel economy required of passenger cars remained
31 percent higher than the average fuel economy required of light
trucks. This methodology resulted in the average fuel economy levels
for passenger cars and light trucks during MYs 2012-2016 as shown in
Table I.B.1-1. The following chart illustrates this methodology of
shifting the standards from the levels maximizing net benefits to the
levels consistent with the combined fuel economy standards in this
final rule.
BILLING CODE 6560-50-P

[[Page 25371]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.013

BILLING CODE 6560-50-C
The final car and truck standards for EPA (Table I.B.1-4 above)
were subsequently determined by first converting the average required
fuel economy levels to average required CO2 emission rates,
and then applying the expected air conditioning credits for 2012-2016.
These A/C credits are shown in the following table. Further details of
the derivation of these factors can be found in Section III of this
preamble or in the EPA RIA.
---------------------------------------------------------------------------

\80\ We assume slightly higher A/C penetration in 2012 than was
assumed in the proposal only to correct for rounding that occurred
in the curve setting process.

[[Page 25372]]

Table II.D-1 Expected Fleet A/C Credits (in CO2 Equivalent g/mi) From 2012-2016
----------------------------------------------------------------------------------------------------------------
Average
technology Average credit Average credit Average credit
penetration for cars for trucks for combined
(%) fleet
----------------------------------------------------------------------------------------------------------------
2012............................................ \80\ 28 3.4 3.8 3.5
2013............................................ 40 4.8 5.4 5.0
2014............................................ 60 7.2 8.1 7.5
2015............................................ 80 9.6 10.8 10.0
2016............................................ 85 10.2 11.5 10.6
----------------------------------------------------------------------------------------------------------------

The agencies sought comment on the use of this methodology for
apportioning the fleet stringencies to relative car and truck standards
for 2012-2016. General Motors commented that, compared to the passenger
car standard, the light truck standard is too stringent because ``the
most fuel efficient cars and small trucks already meet the 2016 MY
requirements'' but ``the most fuel efficient large trucks must increase
fuel economy by 20 percent to meet the 2016 MY requirements.'' GM
recommended that the agencies relax stringency specifically for large
pickups, such as the Silverado.
The agencies disagree with the premise of the comment that the
standard is too stringent under the applicable statutory provisions
because some existing large trucks are not already meeting a later
model year standard. Our analysis shows that the standards are not too
stringent for manufacturers selling these vehicles. The agencies'
analyses demonstrate a means by which manufacturers could apply cost-
effective technologies in order to achieve the standards, and we have
provided adequate lead time for the technology to be applied. More
important, the agencies' analysis demonstrate that the fleetwide
emission standards for MY 2016 are technically feasible, for example by
implementing technologies such as engine downsizing, turbocharging,
direct injection, improving accessories and tire rolling resistance,
etc.
GM did not comment on the use of the methodology applied by the
agencies to develop the gap between the passenger car and light truck
standards--only on the outcome of the methodology. For the reasons
discussed below, the agencies maintain that the methodology applied
above provides an appropriate basis to determine the gap between the
passenger car and light truck standards, and disagree with GM's
arguments that the outcome is unfair.
First, GM's argument incorrectly suggests that every individual
vehicle model must achieve its fuel economy and emissions targets. CAFE
standards and new GHG emissions standards apply to fleetwide average
performance, not model-specific performance, even though average
required levels are based on average model-specific targets, and the
agencies' analysis demonstrates that GM and other manufacturers of
large trucks can cost-effectively comply with the new standards.
Second, GM implies that every manufacturer must be challenged
equally with respect to fuel economy and emissions. Although NHTSA and
EPA maintain that attribute-based CAFE and GHG emissions standards can
more evenly balance compliance challenges, attribute-based standards
are not intended to and cannot make these challenges equal, and while
the agencies are mindful of the potential impacts of the standards on
the relative competitiveness of different vehicle manufacturers, there
is nothing in EPCA or the CAA \81\ requiring that these challenges be
equal.
---------------------------------------------------------------------------

\81\ As NHTSA explained in the NPRM, the Conference Report for
EPCA, as enacted in 1975, makes clear, and the case law affirms, ``a
determination of maximum feasible average fuel economy should not be
keyed to the single manufacturer which might have the most
difficulty achieving a given level of average fuel economy.'' CEI-I,
793 F.2d 1322, 1352 (D.C. Cir. 1986). Instead, NHTSA is compelled
``to weigh the benefits to the nation of a higher fuel economy
standard against the difficulties of individual automobile
manufacturers.'' Id. The law permits CAFE standards exceeding the
projected capability of any particular manufacturer as long as the
standard is economically practicable for the industry as a whole.
Similarly, EPA is afforded great discretion under section 202(a) of
the CAA to balance issues of technical feasibility, cost, adequacy
of lead time, and safety, and certainly is not required to do so in
a manner that imposes regulatory obligations uniformly on each
manufacturer. See NRDC v. EPA, 655 F. 2d 318, 322, 328 (D.C. Cir.
1981) (wide discretion afforded by the statutory factors, and EPA
predictions of technical feasibility afforded considerable
discretion subject to constraints of reasonableness EPA predictions
of technical feasibility afforded considerable discretion subject to
constraints of reasonableness); and cf. International Harvester Co.
v. Ruckelshaus, 479 F. 2d 615, 640 (D.C. Cir. 1973) (``as long as
feasible technology permits the demand for new passenger automobiles
to be generally met, the basic requirements of the Act would be
satisfied, even though this might occasion fewer models and a more
limited choice of engine types'').
---------------------------------------------------------------------------

We have also already addressed and rejected GM's suggestion of
shifting the ``cut off'' point for light trucks from 66 square feet to
72 square feet, thereby ``dropping the floor'' of the target function
for light trucks. As discussed in the preceding section, this is so as
not to forego the rules' energy and environmental benefits, and because
there is little or no safety basis to discourage downsizing of the
largest light trucks.
Finally, NHTSA and EPA disagree with GM's claim that the outcome of
the agencies' approach is unfairly burdensome for light trucks as
compared to passenger cars. Based on the agencies' market forecast,
NHTSA's analysis indicates that incremental technology outlays could,
on average, be comparable for passenger cars and light trucks under the
final CAFE standards, and further indicates that the ratio of total
benefits to total costs could be greater under the final light truck
standards than under the final passenger car standards.

E. Joint Vehicle Technology Assumptions

Vehicle technology assumptions, i.e., assumptions about
technologies' cost, effectiveness, and the rate at which they can be
incorporated into new vehicles, are often controversial as they have a
significant impact on the levels of the standards. The agencies must,
therefore, take great care in developing and justifying these
estimates. In developing technology inputs for the analysis of the MY
2012-2016 standards, the agencies reviewed the technology assumptions
that NHTSA used in setting the MY 2011 standards, the comments that
NHTSA received in response to its May 2008 Notice of Proposed
Rulemaking (NPRM), and the comments received in response to the NPRM
for this rule. This review is consistent with the request by President
Obama in his January 26 memorandum to DOT. In addition, the agencies
reviewed the technology input

[[Page 25373]]

estimates identified in EPA's July 2008 Advance Notice of Proposed
Rulemaking. The review of these documents was supplemented with updated
information from more current literature, new product plans from
manufacturers, and from EPA certification testing.
As a general matter, EPA and NHTSA believe that the best way to
derive technology cost estimates is to conduct real-world tear down
studies. Most of the commenters on this issue agreed. The advantages
not only lie in the rigor of the approach, but also in its
transparency. These studies break down each technology into its
respective components, evaluate the costs of each component, and build
up the costs of the entire technology based on the contribution of each
component and the processes required to integrate them. As such, tear
down studies require a significant amount of time and are very costly.
EPA has been conducting tear down studies to assess the costs of
vehicle technologies under a contract with FEV. Further details for
this methodology is described below and in the TSD.
Due to the complexity and time incurred in a tear down study, only
a few technologies evaluated in this rulemaking have been costed in
this manner thus far. The agencies prioritized the technologies to be
costed first based on how prevalent the agencies believed they might be
likely to be during the rulemaking time frame, and based on their
anticipated cost-effectiveness. The agencies believe that the focus on
these important technologies (listed below) is sufficient for the
analysis in this rule, but EPA is continuing to analyze more
technologies beyond this rule as part of studies both already underway
and in the future. For most of the other technologies, because tear
down studies were not yet available, the agencies decided to pursue, to
the extent possible, the Bill of Materials (BOM) approach as outlined
in NHTSA's MY 2011 final rule. A similar approach was used by EPA in
the EPA 2008 Staff Technical Report. This approach was recommended to
NHTSA by Ricardo, an international engineering consulting firm retained
by NHTSA to aid in the analysis of public comments on its proposed
standards for MYs 2011-2015 because of its expertise in the area of
fuel economy technologies. A BOM approach is one element of the process
used in tear down studies. The difference is that under a BOM approach,
the build up of cost estimates is conducted based on a review of cost
and effectiveness estimates for each component from available
literature, while under a tear down study, the cost estimates which go
into the BOM come from the tear down study itself. To the extent that
the agencies departed from the MY 2011 CAFE final rule estimates, the
agencies explained the reasons and provided supporting analyses in the
Technical Support Document.
Similarly, the agencies followed a BOM approach for developing the
technology effectiveness estimates, insofar as the BOM developed for
the cost estimates helped to inform the appropriate effectiveness
values derived from the literature review. The agencies supplemented
the information with results from available simulation work and real
world EPA certification testing.
The agencies would also like to note that per the Energy
Independence and Security Act (EISA), the National Academies of
Sciences has been conducting a study for NHTSA to update Chapter 3 of
their 2002 NAS Report, which presents technology effectiveness
estimates for light-duty vehicles. The update takes a fresh look at
that list of technologies and their associated cost and effectiveness
values. The updated NAS report was expected to be available on
September 30, 2009, but has not been completed and released to the
public. The results from this study thus are unavailable for this
rulemaking. The agencies look forward to considering the results from
this study as part of the next round of rulemaking for CAFE/GHG
standards.
1. What technologies did the agencies consider?
The agencies considered over 35 vehicle technologies that
manufacturers could use to improve the fuel economy and reduce
CO2 emissions of their vehicles during MYs 2012-2016. The
majority of the technologies described in this section are readily
available, well known, and could be incorporated into vehicles once
production decisions are made. Other technologies considered may not
currently be in production, but are beyond the research phase and under
development, and are expected to be in production in the next few
years. These are technologies which can, for the most part, be applied
both to cars and trucks, and which are capable of achieving significant
improvements in fuel economy and reductions in CO2
emissions, at reasonable costs. The agencies did not consider
technologies in the research stage because the lead time available for
this rule is not sufficient to move most of these technologies from
research to production.
The technologies considered in the agencies' analysis are briefly
described below. They fall into five broad categories: Engine
technologies, transmission technologies, vehicle technologies,
electrification/accessory technologies, and hybrid technologies. For a
more detailed description of each technology and their costs and
effectiveness, we refer the reader to Chapter 3 of the Joint TSD,
Chapter III of NHTSA's FRIA, and Chapter 1 of EPA's final RIA.
Technologies to reduce CO2 and HFC emissions from air
conditioning systems are discussed in Section III of this preamble and
in EPA's final RIA.
Types of engine technologies that improve fuel economy and reduce
CO2 emissions include the following:
Low-friction lubricants--low viscosity and advanced low
friction lubricants oils are now available with improved performance
and better lubrication. If manufacturers choose to make use of these
lubricants, they would need to make engine changes and possibly conduct
durability testing to accommodate the low-friction lubricants.
Reduction of engine friction losses--can be achieved
through low-tension piston rings, roller cam followers, improved
material coatings, more optimal thermal management, piston surface
treatments, and other improvements in the design of engine components
and subsystems that improve engine operation.
Conversion to dual overhead cam with dual cam phasing--as
applied to overhead valves designed to increase the air flow with more
than two valves per cylinder and reduce pumping losses.
Cylinder deactivation--deactivates the intake and exhaust
valves and prevents fuel injection into some cylinders during light-
load operation. The engine runs temporarily as though it were a smaller
engine which substantially reduces pumping losses.
Variable valve timing--alters the timing of the intake
valve, exhaust valve, or both, primarily to reduce pumping losses,
increase specific power, and control residual gases.
Discrete variable valve lift--increases efficiency by
optimizing air flow over a broader range of engine operation which
reduces pumping losses. Accomplished by controlled switching between
two or more cam profile lobe heights.
Continuous variable valve lift--is an electromechanically
controlled system in which valve timing is changed as lift height is
controlled. This yields a wide range of performance

[[Page 25374]]

optimization and volumetric efficiency, including enabling the engine
to be valve throttled.
Stoichiometric gasoline direct-injection technology--
injects fuel at high pressure directly into the combustion chamber to
improve cooling of the air/fuel charge within the cylinder, which
allows for higher compression ratios and increased thermodynamic
efficiency.
Combustion restart--can be used in conjunction with
gasoline direct-injection systems to enable idle-off or start-stop
functionality. Similar to other start-stop technologies, additional
enablers, such as electric power steering, accessory drive components,
and auxiliary oil pump, might be required.
Turbocharging and downsizing--increases the available
airflow and specific power level, allowing a reduced engine size while
maintaining performance. This reduces pumping losses at lighter loads
in comparison to a larger engine.
Exhaust-gas recirculation boost--increases the exhaust-gas
recirculation used in the combustion process to increase thermal
efficiency and reduce pumping losses.
Diesel engines--have several characteristics that give
superior fuel efficiency, including reduced pumping losses due to lack
of (or greatly reduced) throttling, and a combustion cycle that
operates at a higher compression ratio, with a very lean air/fuel
mixture, relative to an equivalent-performance gasoline engine. This
technology requires additional enablers, such as NOX trap
catalyst after-treatment or selective catalytic reduction
NOX after-treatment. The cost and effectiveness estimates
for the diesel engine and aftertreatment system utilized in this final
rule have been revised from the NHTSA MY 2011 CAFE final rule.
Additionally, the diesel technology option has been made available to
small cars in the Volpe and OMEGA models. Though this is not expected
to make a significant difference in the modeling results, the agencies
agreed with the commenters that supported such a revision.
Types of transmission technologies considered include:
Improved automatic transmission controls-- optimizes shift
schedule to maximize fuel efficiency under wide ranging conditions, and
minimizes losses associated with torque converter slip through lock-up
or modulation.
Six-, seven-, and eight-speed automatic transmissions--the
gear ratio spacing and transmission ratio are optimized to enable the
engine to operate in a more efficient operating range over a broader
range of vehicle operating conditions.
Dual clutch or automated shift manual transmissions--are
similar to manual transmissions, but the vehicle controls shifting and
launch functions. A dual-clutch automated shift manual transmission
uses separate clutches for even-numbered and odd-numbered gears, so the
next expected gear is pre-selected, which allows for faster and
smoother shifting.
Continuously variable transmission--commonly uses V-shaped
pulleys connected by a metal belt rather than gears to provide ratios
for operation. Unlike manual and automatic transmissions with fixed
transmission ratios, continuously variable transmissions can provide
fully variable and an infinite number of transmission ratios that
enable the engine to operate in a more efficient operating range over a
broader range of vehicle operating conditions.
Manual 6-speed transmission--offers an additional gear
ratio, often with a higher overdrive gear ratio, than a 5-speed manual
transmission.
Types of vehicle technologies considered include:
Low-rolling-resistance tires--have characteristics that
reduce frictional losses associated with the energy dissipated in the
deformation of the tires under load, thereby improving fuel economy and
reducing CO2 emissions.
Low-drag brakes--reduce the sliding friction of disc brake
pads on rotors when the brakes are not engaged because the brake pads
are pulled away from the rotors.
Front or secondary axle disconnect for four-wheel drive
systems--provides a torque distribution disconnect between front and
rear axles when torque is not required for the non-driving axle. This
results in the reduction of associated parasitic energy losses.
Aerodynamic drag reduction--is achieved by changing
vehicle shape or reducing frontal area, including skirts, air dams,
underbody covers, and more aerodynamic side view mirrors.
Mass reduction and material substitution--Mass reduction
encompasses a variety of techniques ranging from improved design and
better component integration to application of lighter and higher-
strength materials. Mass reduction is further compounded by reductions
in engine power and ancillary systems (transmission, steering, brakes,
suspension, etc.). The agencies recognize there is a range of diversity
and complexity for mass reduction and material substitution
technologies and there are many techniques that automotive suppliers
and manufacturers are using to achieve the levels of this technology
that the agencies have modeled in our analysis for the final standards.
Types of electrification/accessory and hybrid technologies
considered include:
Electric power steering (EPS)--is an electrically-assisted
steering system that has advantages over traditional hydraulic power
steering because it replaces a continuously operated hydraulic pump,
thereby reducing parasitic losses from the accessory drive.
Improved accessories (IACC)--may include high efficiency
alternators, electrically driven (i.e., on-demand) water pumps and
cooling fans. This excludes other electrical accessories such as
electric oil pumps and electrically driven air conditioner compressors.
The latter is covered explicitly within the A/C credit program.
Air Conditioner Systems--These technologies include
improved hoses, connectors and seals for leakage control. They also
include improved compressors, expansion valves, heat exchangers and the
control of these components for the purposes of improving tailpipe
CO2 emissions as a result of A/C use. These technologies are
discussed later in this preamble and covered separately in the EPA RIA.
12-volt micro-hybrid (MHEV)--also known as idle-stop or
start-stop and commonly implemented as a 12-volt belt-driven integrated
starter-generator, this is the most basic hybrid system that
facilitates idle-stop capability. Along with other enablers, this
system replaces a common alternator with a belt-driven enhanced power
starter-alternator, and a revised accessory drive system.
Higher Voltage Stop-Start/Belt Integrated Starter
Generator (BISG)--provides idle-stop capability and uses a higher
voltage battery with increased energy capacity over typical automotive
batteries. The higher system voltage allows the use of a smaller, more
powerful electric motor. This system replaces a standard alternator
with an enhanced power, higher voltage, higher efficiency starter-
alternator, that is belt driven and that can recover braking energy
while the vehicle slows down (regenerative braking).
Integrated Motor Assist (IMA)/Crank integrated starter
generator (CISG)--provides idle-stop capability and uses a high voltage
battery with increased energy capacity over typical automotive
batteries. The higher system voltage allows the use of a smaller, more

[[Page 25375]]

powerful electric motor and reduces the weight of the wiring harness.
This system replaces a standard alternator with an enhanced power,
higher voltage, higher efficiency starter-alternator that is crankshaft
mounted and can recover braking energy while the vehicle slows down
(regenerative braking).
2-mode hybrid (2MHEV)--is a hybrid electric drive system
that uses an adaptation of a conventional stepped-ratio automatic
transmission by replacing some of the transmission clutches with two
electric motors that control the ratio of engine speed to vehicle
speed, while clutches allow the motors to be bypassed. This improves
both the transmission torque capacity for heavy-duty applications and
reduces fuel consumption and CO2 emissions at highway speeds
relative to other types of hybrid electric drive systems.
Power-split hybrid (PSHEV)--a hybrid electric drive system
that replaces the traditional transmission with a single planetary
gearset and a motor/generator. This motor/generator uses the engine to
either charge the battery or supply additional power to the drive
motor. A second, more powerful motor/generator is permanently connected
to the vehicle's final drive and always turns with the wheels. The
planetary gear splits engine power between the first motor/generator
and the drive motor to either charge the battery or supply power to the
wheels.
Plug-in hybrid electric vehicles (PHEV)--are hybrid
electric vehicles with the means to charge their battery packs from an
outside source of electricity (usually the electric grid). These
vehicles have larger battery packs with more energy storage and a
greater capability to be discharged than other hybrids. They also use a
control system that allows the battery pack to be substantially
depleted under electric-only or blended mechanical/electric operation.
Electric vehicles (EV)--are vehicles with all-electric
drive and with vehicle systems powered by energy-optimized batteries
charged primarily from grid electricity.
The cost estimates for the various hybrid systems have been revised
from the estimates used in the MY 2011 CAFE final rule, in particular
with respect to estimated battery costs.
2. How did the agencies determine the costs and effectiveness of each
of these technologies?
As mentioned above, EPA and NHTSA believe that the best way to
derive technology cost estimates is to conduct real-world tear down
studies. To date, the costs of the following five technologies have
been evaluated with respect to their baseline (or replaced)
technologies. For these technologies noted below, the agencies relied
on the tear down data available and scaling methodologies used in EPA's
ongoing study with FEV. Only the cost estimate for the first technology
on the list below was used in the NPRM. The others were completed
subsequent to the publication of the NPRM.
1. Stoichiometric gasoline direct injection and turbo charging with
engine downsizing (T-DS) for a large DOHC 4 cylinder engine to a small
DOHC (dual overhead cam) 4 cylinder engine.
2. Stoichiometric gasoline direct injection and turbo charging with
engine downsizing for a SOHC single overhead cam) 3 valve/cylinder V8
engine to a SOHC V6 engine.
3. Stoichiometric gasoline direct injection and turbo charging with
engine downsizing for a DOHC V6 engine to a DOHC 4 cylinder engine.
4. 6-speed automatic transmission replacing a 5-speed automatic
transmission.
5. 6-speed wet dual clutch transmission (DCT) replacing a 6-speed
automatic transmission.
This costing methodology has been published and gone through a peer
review.\82\ Using this tear down costing methodology, FEV has developed
costs for each of the above technologies. In addition, FEV and EPA
extrapolated the engine downsizing costs for the following scenarios
that were outside of the noted study cases:\83\
---------------------------------------------------------------------------

\82\ EPA-420-R-09-020; EPA docket number EPA-HQ-OAR-2009-0472-
11282 and 11285.
\83\ ``Binning of FEV Costs to GDI, Turbo-charging, and Engine
Downsizing,'' memorandum to Docket EPA-HQ-OAR-2009-0472, from
Michael Olechiw, U.S. EPA, dated March 25, 2010.
---------------------------------------------------------------------------

1. Downsizing a SOHC 2 valve/cylinder V8 engine to a DOHC V6.
2. Downsizing a DOHC V8 to a DOHC V6.
3. Downsizing a SOHC V6 engine to a DOHC 4 cylinder engine.
4. Downsizing a DOHC 4 cylinder engine to a DOHC 3 cylinder engine.
The agencies relied on the findings of FEV in part for estimating
the cost of these technologies in this rulemaking. However, for some of
the technologies, NHTSA and EPA modified FEV's estimated costs. FEV
made the assumption that these technologies would be mature when
produced in large volumes (450,000 units or more). The agencies believe
that there is some uncertainty regarding each manufacturer's near-term
ability to employ the technology at the volumes assumed in the FEV
analysis. There is also the potential for near term (earlier than 2016)
supplier-level Engineering, Design and Testing (ED&T) costs to be in
excess of those considered in the FEV analysis as existing equipment
and facilities are converted to production of new technologies. The
agencies have therefore decided to average the FEV results with the
NPRM values in an effort to account for these near-term factors. This
methodology was done for the following technologies:
1. Converting a port-fuel injected (PFI) DOHC I4 to a turbocharged-
downsized-stoichiometric GDI DOHC I3.
2. Converting a PFI DOHC V6 engine to a T-DS-stoichiometric GDI
DOHC I4.
3. Converting a PFI SOHC V6 engine to a T-DS-stoichiometric GDI
DOHC I4.
4. Converting a PFI DOHC V8 engine to a T-DS-stoichiometric GDI
DOHC V6.
5. Converting a PFI SOHC 3V V8 engine to a T-DS-stoichiometric GDI
DOHC V6.
6. Converting a PFI SOHC 2V V8 engine to a T-DS-stoichiometric GDI
DOHC V6.
7. Replacing a 4-speed automatic transmission with a 6-speed
automatic transmission.
8. Replacing a 5-speed automatic transmission with a 6-speed
automatic transmission.
9. Replacing a 6-speed automatic transmission with a 6-speed wet
dual clutch transmission.
For the I4 to Turbo GDI I4 study applied in the NPRM, the agencies
requested from FEV an adjusted cost estimate which accounted for these
uncertainties as an adjustment to the base technology burden rate.\84\
These new costs are used in the final rules. These details are also
further described in the memo to the docket.\85\ The confidential
information provided by manufacturers as part of their product plan
submissions to the agencies or discussed in meetings between the
agencies and the manufacturers and

[[Page 25376]]

suppliers served largely as a check on publicly-available data.
---------------------------------------------------------------------------

\84\ Burden costs include the following fixed and variable
costs: Rented and leased equipment; manufacturing equipment
depreciation; plant office equipment depreciation; utilities
expense; insurance (fire and general); municipal taxes; plant floor
space (equipment and plant offices); maintenance of manufacturing
equipment--non-labor; maintenance of manufacturing building--
general, internal and external, parts, and labor; operating
supplies; perishable and supplier-owned tooling; all other plant
wages (excluding direct, indirect and MRO labor); returnable dunnage
maintenance; and intra-company shipping costs (see EPA-HQ-OAR-2009-
0472-0149).
\85\ ``Binning of FEV Costs to GDI, Turbo-charging, and Engine
Downsizing,'' memorandum to Docket EPA-HQ-OAR-2009-0472, from
Michael Olechiw, U.S. EPA, dated March 25, 2010.
---------------------------------------------------------------------------

For the other technologies, considering all sources of information
(including public comments) and using the BOM approach, the agencies
worked together intensively to determine component costs for each of
the technologies and build up the costs accordingly. Where estimates
differ between sources, we have used our engineering judgment to arrive
at what we believe to be the best available cost estimate, and
explained the basis for that exercise of judgment in the TSD. Building
on NHTSA's estimates developed for the MY 2011 CAFE final rule and
EPA's Advance Notice of Proposed Rulemaking, which relied on the EPA
2008 Staff Technical Report,\86\ the agencies took a fresh look at
technology cost and effectiveness values for purposes of the joint
rulemaking under the National Program. For costs, the agencies
reconsidered both the direct or ``piece'' costs and indirect costs of
individual components of technologies. For the direct costs, the
agencies followed a bill of materials (BOM) approach employed in
NHTSA's MY 2011 final rule based on recommendation from Ricardo, Inc.,
as described above. EPA used a similar approach in the EPA 2008 Staff
Technical Report. A bill of materials, in a general sense, is a list of
components or sub-systems that make up a system--in this case, an item
of fuel economy-improving technology. In order to determine what a
system costs, one of the first steps is to determine its components and
what they cost.
---------------------------------------------------------------------------

\86\ EPA Staff Technical Report: Cost and Effectiveness
Estimates of Technologies Used to Reduce Light-Duty Vehicle Carbon
Dioxide Emissions. EPA420-R-08-008, March 2008.
---------------------------------------------------------------------------

NHTSA and EPA estimated these components and their costs based on a
number of sources for cost-related information. The objective was to
use those sources of information considered to be most credible for
projecting the costs of individual vehicle technologies. For example,
while NHTSA and Ricardo engineers had relied considerably in the MY
2011 final rule on the 2008 Martec Report for costing contents of some
technologies, upon further joint review and for purposes of the MY
2012-2016 standards, the agencies decided that some of the costing
information in that report was no longer accurate due to downward
trends in commodity prices since the publication of that report. The
agencies reviewed, then revalidated or updated cost estimates for
individual components based on new information. Thus, while NHTSA and
EPA found that much of the cost information used in NHTSA's MY 2011
final rule and EPA's staff report was consistent to a great extent, the
agencies, in reconsidering information from many
sources,^{87 88 89 90 91 92 93} revised several component costs
of several major technologies: turbocharging with engine downsizing (as
described above), mild and strong hybrids, diesels, stoichiometric
gasoline direct injection fuel systems, and valve train lift
technologies. These are discussed at length in the Joint TSD and in
NHTSA's final RIA.
---------------------------------------------------------------------------

\87\ National Research Council, ``Effectiveness and Impact of
Corporate Average Fuel Economy (CAFE) Standards,'' National Academy
Press, Washington, DC (2002) (the ``2002 NAS Report''), available at
http://www.nap.edu/openbook.php?isbn=0309076013 (last accessed
August 7, 2009--update).
\88\ Northeast States Center for a Clean Air Future (NESCCAF),
``Reducing Greenhouse Gas Emissions from Light-Duty Motor
Vehicles,'' 2004 (the ``2004 NESCCAF Report''), available at http://
www.nesccaf.org/documents/rpt040923ghglightduty.pdf (last accessed
August 7, 2009--update).
\89\ ``Staff Report: Initial Statement of Reasons for Proposed
Rulemaking, Public Hearing to Consider Adoption of Regulations to
Control Greenhouse Gas Emissions from Motor Vehicles,'' California
Environmental Protection Agency, Air Resources Board, August 6,
2004.
\90\ Energy and Environmental Analysis, Inc., ``Technology to
Improve the Fuel Economy of Light Duty Trucks to 2015,'' 2006 (the
``2006 EEA Report''), Docket EPA-HQ-OAR-2009-0472.
\91\ Martec, ``Variable Costs of Fuel Economy Technologies,''
June 1, 2008, (the ``2008 Martec Report'') available at Docket No.
NHTSA-2008-0089-0169.1.
\92\ Vehicle fuel economy certification data.
\93\ Confidential data submitted by manufacturers in response to
the March 2009 and other requests for product plans.
---------------------------------------------------------------------------

Once costs were determined, they were adjusted to ensure that they
were all expressed in 2007 dollars using a ratio of GDP values for the
associated calendar years,\94\ and indirect costs were accounted for
using the ICM (indirect cost multiplier) approach explained in Chapter
3 of the Joint TSD, rather than using the traditional Retail Price
Equivalent (RPE) multiplier approach. A report explaining how EPA
developed the ICM approach can be found in the docket for this rule.
The comments addressing the ICM approach were generally positive and
encouraging. However, one commenter suggested that we had
mischaracterized the complexity of a few of our technologies, which
would result in higher or lower markups than presented in the NPRM.
That commenter also suggested that we had used the ICMs as a means of
placing a higher level of manufacturer learning on the cost estimates.
The latter comment is not true and the methodology behind the ICM
approach is explained in detail in the reports that are available in
the docket for this rule.\95\ The former is open to debate given the
subjective nature of the engineering analysis behind it, but upon
further thought both agencies believe that the complexities used in the
NPRM were appropriate and have, therefore, carried those forward into
the final rule. We discuss this in greater detail in the Response to
Comments document.
---------------------------------------------------------------------------

\94\ NHTSA examined the use of the CPI multiplier instead of GDP
for adjusting these dollar values, but found the difference to be
exceedingly small--only $0.14 over $100.
\95\ Rogozhin, Alex, Michael Gallaher, and Walter McManus,
``Automobile Industry Retail Price Equivalent and Indirect Cost
Multipliers,'' EPA 420-R-09-003, Docket EPA Docket EPA-HQ-OAR-2009-
0472-0142, February 2009, http://epa.gov/otaq/ld-hwy/420r09003.pdf;
A. Rogozhin et al., International Journal of Production Economics
124 (2010) 360-368, Volume 124, Issue 2, April 2010.
---------------------------------------------------------------------------

Regarding estimates for technology effectiveness, NHTSA and EPA
also reexamined the estimates from NHTSA's MY 2011 final rule and EPA's
ANPRM and 2008 Staff Technical Report, which were largely consistent
with NHTSA's 2008 NPRM estimates. The agencies also reconsidered other
sources such as the 2002 NAS Report, the 2004 NESCCAF report, recent
CAFE compliance data (by comparing similar vehicles with different
technologies against each other in fuel economy testing, such as a
Honda Civic Hybrid versus a directly comparable Honda Civic
conventional drive), and confidential manufacturer estimates of
technology effectiveness. NHTSA and EPA engineers reviewed
effectiveness information from the multiple sources for each technology
and ensured that such effectiveness estimates were based on technology
hardware consistent with the BOM components used to estimate costs. The
agencies also carefully examined the pertinent public comments.
Together, they compared the multiple estimates and assessed their
validity, taking care to ensure that common BOM definitions and other
vehicle attributes such as performance, refinement, and drivability
were taken into account. However, because the agencies' respective
models employ different numbers of vehicle subclasses and use different
modeling techniques to arrive at the standards, direct comparison of
BOMs was somewhat more complicated. To address this and to confirm that
the outputs from the different modeling techniques produced the same
result, NHTSA and EPA developed mapping techniques, devising technology
packages and mapping them to corresponding incremental technology
estimates. This approach helped compare the outputs

[[Page 25377]]

from the incremental modeling technique to those produced by the
technology packaging approach to ensure results that are consistent and
could be translated into the respective models of the agencies.
In general, most effectiveness estimates used in both the MY 2011
final rule and the 2008 EPA staff report were determined to be accurate
and were carried forward without significant change first into the
NPRM, and now into these final rules. When NHTSA and EPA's estimates
for effectiveness diverged slightly due to differences in how the
agencies apply technologies to vehicles in their respective models, we
report the ranges for the effectiveness values used in each model.
There were only a few comments on the technology effectiveness
estimates used in the NPRM. Most of the technologies that were
mentioned in the comments were the more advanced technologies that are
not assumed to have large penetrations in the market within the
timeframe of this rule, notably hybrid technologies. Even if the
effectiveness figures for hybrid vehicles were adjusted, it would have
made little difference in the NHTSA and EPA analysis of the impacts and
costs of the rule. The response to comments document has more specific
responses to these comments.
The agencies note that the effectiveness values estimated for the
technologies considered in the modeling analyses may represent average
values, and do not reflect the enormous spectrum of possible values
that could result from adding the technology to different vehicles. For
example, while the agencies have estimated an effectiveness of 0.5
percent for low friction lubricants, each vehicle could have a unique
effectiveness estimate depending on the baseline vehicle's oil
viscosity rating. Similarly, the reduction in rolling resistance (and
thus the improvement in fuel economy and the reduction in
CO2 emissions) due to the application of low rolling
resistance tires depends not only on the unique characteristics of the
tires originally on the vehicle, but on the unique characteristics of
the tires being applied, characteristics which must be balanced between
fuel efficiency, safety, and performance. Aerodynamic drag reduction is
much the same--it can improve fuel economy and reduce CO2
emissions, but it is also highly dependent on vehicle-specific
functional objectives. For purposes of the final standards, NHTSA and
EPA believe that employing average values for technology effectiveness
estimates, as adjusted depending on vehicle subclass, is an appropriate
way of recognizing the potential variation in the specific benefits
that individual manufacturers (and individual vehicles) might obtain
from adding a fuel-saving technology.
Chapter 3 of the Joint Technical Support Document contains a
detailed description of our assessment of vehicle technology cost and
effectiveness estimates. The agencies note that the technology costs
included in this final rule take into account only those associated
with the initial build of the vehicle. Although comments were received
to the NPRM that suggested there could be additional maintenance
required with some new technologies (e.g., turbocharging, hybrids,
etc.), and that additional maintenance costs could occur as a result,
the agencies do not believe that the amount of additional cost will be
significant in the timeframe of this rulemaking, based on the
relatively low application rates for these technologies. The agencies
will undertake a more detailed review of these potential costs in
preparation for the next round of CAFE/GHG standards.

F. Joint Economic Assumptions

The agencies' final analysis of alternative CAFE and GHG standards
for the model years covered by this final rulemaking rely on a range of
forecast information, economic estimates, and input parameters. This
section briefly describes the agencies' choices of specific parameter
values. These economic values play a significant role in determining
the benefits of both CAFE and GHG standards.
In reviewing these variables and the agency's estimates of their
values for purposes of this final rule, NHTSA and EPA reconsidered
previous comments that NHTSA had received, reviewed newly available
literature, and reviewed comments received in response to the proposed
rule. For this final rule, we made three major changes to the economic
assumptions. First, we revised the technology costs to reflect more
recently available data. Second, we updated fuel price and
transportation demand assumptions to reflect the Annual Energy Outlook
(AEO) 2010 Early Release. Third, we have updated our estimates of the
social cost of carbon (SCC) based on a recent interagency process. The
key economic assumptions are summarized below, and are discussed in
greater detail in Section III (EPA) and Section IV (NHTSA), as well as
in Chapter 4 of the Joint TSD, Chapter VIII of NHTSA's RIA and Chapter
8 of EPA's RIA.
Costs of fuel economy-improving technologies--These
estimates are presented in summary form above and in more detail in the
agencies' respective sections of this preamble, in Chapter 3 of the
Joint TSD, and in the agencies' respective RIAs. The technology cost
estimates used in this analysis are intended to represent
manufacturers' direct costs for high-volume production of vehicles with
these technologies and sufficient experience with their application so
that all cost reductions due to ``learning curve'' effects have been
fully realized. Costs are then modified by applying near-term indirect
cost multipliers ranging from 1.11 to 1.64 to the estimates of vehicle
manufacturers' direct costs for producing or acquiring each technology
to improve fuel economy, depending on the complexity of the technology
and the time frame over which costs are estimated. This accounts for
both the direct and indirect costs associated with implementing new
technologies in response to this final rule. The technology cost
estimates for a select group of technologies have changed since the
NPRM. These changes, as summarized in Section II.E and in Chapter 3 of
the Joint TSD, were made in response to updated cost estimates
available to the agencies shortly after publication of the NPRM, not in
response to comments. In general, commenters were supportive of the
cost estimates used in the NPRM and the transparency of the methodology
used to generate them.
Potential opportunity costs of improved fuel economy--This
estimate addresses the possibility that achieving the fuel economy
improvements required by alternative CAFE or GHG standards would
require manufacturers to compromise the performance, carrying capacity,
safety, or comfort of their vehicle models. If it did so, the resulting
sacrifice in the value of these attributes to consumers would represent
an additional cost of achieving the required improvements, and thus of
manufacturers' compliance with stricter standards. Currently the
agencies assume that these vehicle attributes do not change, and
include the cost of maintaining these attributes as part of the cost
estimates for technologies. However, it is possible that the technology
cost estimates do not include adequate allowance for the necessary
efforts by manufacturers to maintain vehicle performance, carrying
capacity, and utility while improving fuel economy and reducing GHG
emissions. While, in principle, consumer vehicle demand models can
measure these effects, these models do not appear to be robust across
specifications, since authors derive a

[[Page 25378]]

wide range of willingness-to-pay values for fuel economy from these
models, and there is not clear guidance from the literature on whether
one specification is clearly preferred over another. This issue is
discussed in EPA's RIA, Section 8.1.2 and NHTSA's RIA Section VIII.H.
The agencies requested comment on how to estimate explicitly the
changes in vehicle buyers' welfare from the combination of higher
prices for new vehicle models, increases in their fuel economy, and any
accompanying changes in vehicle attributes such as performance,
passenger- and cargo-carrying capacity, or other dimensions of utility.
Commenters did not provide recommendations for how to evaluate the
quality of different models or identify a model appropriate for the
agencies' purposes. Some commenters expressed various concerns about
the use of existing consumer vehicle choice models. While EPA and NHTSA
are not using a consumer vehicle choice model to analyze the effects of
this rule, we continue to investigate these models.
The on-road fuel economy ``gap''--Actual fuel economy
levels achieved by light-duty vehicles in on-road driving fall somewhat
short of their levels measured under the laboratory-like test
conditions used by NHTSA and EPA to establish compliance with the final
CAFE and GHG standards. The agencies use an on-road fuel economy gap
for light-duty vehicles of 20 percent lower than published fuel economy
levels. For example, if the measured CAFE fuel economy value of a light
truck is 20 mpg, the on-road fuel economy actually achieved by a
typical driver of that vehicle is expected to be 16 mpg (20*.80).\96\
NHTSA previously used this estimate in its MY 2011 final rule, and the
agencies confirmed it based on independent analysis for use in this
FRM. No substantive comments were received on this input.
---------------------------------------------------------------------------

\96\ U.S. Environmental Protection Agency, Final Technical
Support Document, Fuel Economy Labeling of Motor Vehicle Revisions
to Improve Calculation of Fuel Economy Estimates, EPA420-R-06-017,
December 2006.
---------------------------------------------------------------------------

Fuel prices and the value of saving fuel--Projected future
fuel prices are a critical input into the preliminary economic analysis
of alternative standards, because they determine the value of fuel
savings both to new vehicle buyers and to society. For the proposed
rule, the agencies had relied on the then most recent fuel price
projections from the U.S. Energy Information Administration's (EIA)
Annual Energy Outlook (AEO) 2009 (Revised Updated). However, for this
final rule, the agencies have updated the analyses based on AEO 2010
(December 2009 Early Release) Reference Case forecasts of inflation-
adjusted (constant-dollar) retail gasoline and diesel fuel prices,
which represent the EIA's most up-to-date estimate of the most likely
course of future prices for petroleum products.\97\ AEO 2010 includes
slightly lower petroleum prices compared to AEO 2009.
---------------------------------------------------------------------------

\97\ Energy Information Administration, Annual Energy Outlook
2010, Early Release Reference Case (December 2009), Table 12.
Available at http://www.eia.doe.gov/oiaf/aeo/aeoref_tab.html (last
accessed February 02, 2010).
---------------------------------------------------------------------------

The forecasts of fuel prices reported in EIA's AEO 2010 Early
Release Reference Case extends through 2035, compared to the AEO 2009
which only went through 2030. As in the proposal, fuel prices beyond
the time frame of AEO's forecast were estimated using an average growth
rate.
While EIA revised AEO 2010, the vehicle MPG standards are similar
to those that were published in AEO 2009. No substantive comments were
received on the use of AEO as a source of fuel prices.\98\
---------------------------------------------------------------------------

\98\ Kahan, A. and Pickrell, D. Memo to Docket EPA-HQ-OAR-2009-
0472 and Docket NHTSA-2009-0059. ``Energy Information
Administration's Annual Energy Outlook 2009 and 2010.'' March 24,
2010.
---------------------------------------------------------------------------

Consumer valuation of fuel economy and payback period--In
estimating the impacts on vehicle sales, the agencies assume that
potential buyers value the resulting fuel savings improvements that
would result from alternative CAFE and GHG standards over only part of
the expected lifetime of the vehicles they purchase. Specifically, we
assume that buyers value fuel savings over the first five years of a
new vehicle's lifetime, and that buyers discount the value of these
future fuel savings using rates of 3% and 7%. The five-year figure
represents the current average term of consumer loans to finance the
purchase of new vehicles. One commenter argued that higher-fuel-economy
vehicles should have higher resale prices than vehicles with lower fuel
economy, but did not provide supporting data. This revision, if made,
would increase the net benefits of the rule. Another commenter
supported the use of a five-year payback period for this analysis. In
the absence of data to support changes, EPA and NHTSA have kept the
same assumptions. In the analysis of net benefits, EPA and NHTSA assume
that vehicle buyers benefit from the full fuel savings over the
vehicle's lifetime, discounted for present value calculations at 3 and
7 percent.
Vehicle sales assumptions--The first step in estimating
lifetime fuel consumption by vehicles produced during a model year is
to calculate the number of vehicles expected to be produced and
sold.\99\ The agencies relied on the AEO 2010 Early Release for
forecasts of total vehicle sales, while the baseline market forecast
developed by the agencies (see Section II.B) divided total projected
sales into sales of cars and light trucks.
---------------------------------------------------------------------------

\99\ Vehicles are defined to be of age 1 during the calendar
year corresponding to the model year in which they are produced;
thus for example, model year 2000 vehicles are considered to be of
age 1 during calendar year 2000, age 2 during calendar year 2001,
and to reach their maximum age of 26 years during calendar year
2025. NHTSA considers the maximum lifetime of vehicles to be the age
after which less than 2 percent of the vehicles originally produced
during a model year remain in service. Applying these conventions to
vehicle registration data indicates that passenger cars have a
maximum age of 26 years, while light trucks have a maximum lifetime
of 36 years. See Lu, S., NHTSA, Regulatory Analysis and Evaluation
Division, ``Vehicle Survivability and Travel Mileage Schedules,''
DOT HS 809 952, 8-11 (January 2006). Available at http://www-
nrd.nhtsa.dot.gov/Pubs/809952.pdf (last accessed Feb. 15, 2010).
---------------------------------------------------------------------------

Vehicle survival assumptions--We then applied updated
values of age-specific survival rates for cars and light trucks to
these adjusted forecasts of passenger car and light truck sales to
determine the number of these vehicles remaining in use during each
year of their expected lifetimes. No substantive comments were received
on vehicle survival assumptions.
Total vehicle use--We then calculated the total number of
miles that cars and light trucks produced in each model year will be
driven during each year of their lifetimes using estimates of annual
vehicle use by age tabulated from the Federal Highway Administration's
2001 National Household Transportation Survey (NHTS),\100\ adjusted to
account for the effect on vehicle use of subsequent increases in fuel
prices. Due to the lower fuel prices projected in AEO 2010, the average
vehicle is estimated to be used slightly more (~3 percent) over its
lifetime than assumed in the proposal. In order to insure that the
resulting mileage schedules imply reasonable estimates of future growth
in total car and light truck use, we calculated the rate of growth in
annual car and light truck mileage at each age that is necessary for
total car and light truck travel to increase at the rates forecast in
the AEO 2010 Early Release Reference Case. The growth rate in average
annual car and light truck use produced by this calculation is

[[Page 25379]]

approximately 1.1 percent per year.\101\ This rate was applied to the
mileage figures derived from the 2001 NHTS to estimate annual mileage
during each year of the expected lifetimes of MY 2012-2016 cars and
light trucks.\102\ While commenters requested further detail on the
assumptions regarding total vehicle use, no specific issues were
raised.
---------------------------------------------------------------------------

\100\ For a description of the Survey, see http://nhts.ornl.gov/
quickStart.shtml (last accessed July 27, 2009).
\101\ It was not possible to estimate separate growth rates in
average annual use for cars and light trucks, because of the
significant reclassification of light truck models as passenger cars
discussed previously.
\102\ While the adjustment for future fuel prices reduces
average mileage at each age from the values derived from the 2001
NHTS, the adjustment for expected future growth in average vehicle
use increases it. The net effect of these two adjustments is to
increase expected lifetime mileage by about 18 percent for passenger
cars and about 16 percent for light trucks.
---------------------------------------------------------------------------

Accounting for the rebound effect of higher fuel economy--
The rebound effect refers to the fraction of fuel savings expected to
result from an increase in vehicle fuel economy--particularly an
increase required by the adoption of more stringent CAFE and GHG
standards--that is offset by additional vehicle use. The increase in
vehicle use occurs because higher fuel economy reduces the fuel cost of
driving, typically the largest single component of the monetary cost of
operating a vehicle, and vehicle owners respond to this reduction in
operating costs by driving slightly more. We received comments
supporting our proposed value of 10 percent, although we also received
comments recommending higher and lower values. However, we did not
receive any new data or comments that justify revising the 10 percent
value for the rebound effect at this time.
Benefits from increased vehicle use--The increase in
vehicle use from the rebound effect provides additional benefits to
their owners, who may make more frequent trips or travel farther to
reach more desirable destinations. This additional travel provides
benefits to drivers and their passengers by improving their access to
social and economic opportunities away from home. These benefits are
measured by the net ``consumer surplus'' resulting from increased
vehicle use, over and above the fuel expenses associated with this
additional travel. We estimate the economic value of the consumer
surplus provided by added driving using the conventional approximation,
which is one half of the product of the decline in vehicle operating
costs per vehicle-mile and the resulting increase in the annual number
of miles driven. Because it depends on the extent of improvement in
fuel economy, the value of benefits from increased vehicle use changes
by model year and varies among alternative standards.
The value of increased driving range--By reducing the
frequency with which drivers typically refuel their vehicles, and by
extending the upper limit of the range they can travel before requiring
refueling, improving fuel economy and reducing GHG emissions thus
provides some additional benefits to their owners. No direct estimates
of the value of extended vehicle range are readily available, so the
agencies' analysis calculates the reduction in the annual number of
required refueling cycles that results from improved fuel economy, and
applies DOT-recommended values of travel time savings to convert the
resulting time savings to their economic value.\103\ Please see the
Chapter 4 of the Joint TSD for details.
---------------------------------------------------------------------------

\103\ Department of Transportation, Guidance Memorandum, ``The
Value of Saving Travel Time: Departmental Guidance for Conducting
Economic Evaluations,'' Apr. 9, 1997. http://ostpxweb.dot.gov/
policy/Data/VOT97guid.pdf (last accessed Feb. 15, 2010); update
available at http://ostpxweb.dot.gov/policy/Data/VOTrevision1_2-11-
03.pdf (last accessed Feb. 15, 2010).
---------------------------------------------------------------------------

Added costs from congestion, crashes and noise--Although
it provides some benefits to drivers, increased vehicle use associated
with the rebound effect also contributes to increased traffic
congestion, motor vehicle accidents, and highway noise. Depending on
how the additional travel is distributed over the day and on where it
takes place, additional vehicle use can contribute to traffic
congestion and delays by increasing traffic volumes on facilities that
are already heavily traveled during peak periods. These added delays
impose higher costs on drivers and other vehicle occupants in the form
of increased travel time and operating expenses, increased costs
associated with traffic accidents, and increased traffic noise. The
agencies rely on estimates of congestion, accident, and noise costs
caused by automobiles and light trucks developed by the Federal Highway
Administration to estimate the increased external costs caused by added
driving due to the rebound effect.\104\
---------------------------------------------------------------------------

\104\ These estimates were developed by FHWA for use in its 1997
Federal Highway Cost Allocation Study; http://www.fhwa.dot.gov/
policy/hcas/final/index.htm (last accessed Feb. 15, 2010).
---------------------------------------------------------------------------

Petroleum consumption and import externalities--U.S.
consumption and imports of petroleum products also impose costs on the
domestic economy that are not reflected in the market price for crude
petroleum, or in the prices paid by consumers of petroleum products
such as gasoline. In economics literature on this subject, these costs
include (1) higher prices for petroleum products resulting from the
effect of U.S. oil import demand on the world oil price (``monopsony
costs''); (2) the expected costs from the risk of disruptions to the
U.S. economy caused by sudden reductions in the supply of imported oil
to the U.S.; and (3) expenses for maintaining a U.S. military presence
to secure imported oil supplies from unstable regions, and for
maintaining the strategic petroleum reserve (SPR) to cushion against
resulting price increases.\105\ Reducing U.S. imports of crude
petroleum or refined fuels can reduce the magnitude of these external
costs. Any reduction in their total value that results from lower fuel
consumption and petroleum imports represents an economic benefit of
setting more stringent standards over and above the dollar value of
fuel savings itself. Since the agencies are taking a global perspective
with respect to the estimate of the social cost of carbon for this
rulemaking, the agencies do not include the value of any reduction in
monopsony payments as a benefit from lower fuel consumption, because
those payments from a global perspective represent a transfer of income
from consumers of petroleum products to oil suppliers rather than a
savings in real economic resources. Similarly, the agencies do not
include any savings in budgetary outlays to support U.S. military
activities among the benefits of higher fuel economy and the resulting
fuel savings. Based on a recently-updated ORNL study, we estimate that
each gallon of fuel saved that results in a reduction in U.S. petroleum
imports (either crude petroleum or refined fuel) will reduce the
expected costs of oil supply disruptions to the U.S. economy by $0.169
(2007$). Each gallon of fuel saved as a consequence of higher standards
is anticipated to reduce total U.S. imports of crude petroleum or
refined fuel by 0.95 gallons.\106\
---------------------------------------------------------------------------

\105\ See, e.g., Bohi, Douglas R. and W. David Montgomery
(1982). Oil Prices, Energy Security, and Import Policy Washington,
DC: Resources for the Future, Johns Hopkins University Press; Bohi,
D. R., and M. A. Toman (1993). ``Energy and Security: Externalities
and Policies,'' Energy Policy 21:1093-1109; and Toman, M. A. (1993).
``The Economics of Energy Security: Theory, Evidence, Policy,'' in
A. V. Kneese and J. L. Sweeney, eds. (1993). Handbook of Natural
Resource and Energy Economics, Vol. III. Amsterdam: North-Holland,
pp. 1167-1218.
\106\ Each gallon of fuel saved is assumed to reduce imports of
refined fuel by 0.5 gallons, and the volume of fuel refined
domestically by 0.5 gallons. Domestic fuel refining is assumed to
utilize 90 percent imported crude petroleum and 10 percent
domestically-produced crude petroleum as feedstocks. Together, these
assumptions imply that each gallon of fuel saved will reduce imports
of refined fuel and crude petroleum by 0.50 gallons + 0.50
gallons*90 percent = 0.50 gallons + 0.45 gallons = 0.95 gallons.

---------------------------------------------------------------------------

[[Page 25380]]

The energy security analysis conducted for this rule estimates that
the world price of oil will fall modestly in response to lower U.S.
demand for refined fuel. One potential result of this decline in the
world price of oil would be an increase in the consumption of petroleum
products outside the U.S., which would in turn lead to a modest
increase in emissions of greenhouse gases, criteria air pollutants, and
airborne toxics from their refining and use. While additional
information would be needed to analyze this ``leakage effect'' in
detail, NHTSA provides a sample estimate of its potential magnitude in
its Final EIS.\107\ This analysis indicates that the leakage effect is
likely to offset only a modest fraction of the reductions in emissions
projected to result from the rule.
---------------------------------------------------------------------------

\107\ NHTSA Final Environmental Impact Statement: Corporate
Average Fuel Economy Standards, Passenger Cars and Light Trucks,
Model Years 2012-2016, February 2010, page 3-14.
---------------------------------------------------------------------------

EPA and NHTSA received comments about the treatment of the
monopsony effect, macroeconomic disruption effect, and the military
costs associated with the energy security benefits of this rule. The
agencies did not receive any comments that justify changing the energy
security analysis. As a result, the agencies continue to only use the
macroeconomic disruption component of the energy security analysis
under a global context when estimating the total energy security
benefits associated with this rule. Further, the Agencies did not
receive any information that they could use to quantity that component
of military costs directly related to energy security, and thus did not
modify that part of its analysis. A more complete discussion of the
energy security analysis can be found in Chapter 4 of the Joint TSD,
and Sections III and IV of this preamble.
Air pollutant emissions
[cir] Impacts on criteria air pollutant emissions--While reductions
in domestic fuel refining and distribution that result from lower fuel
consumption will reduce U.S. emissions of criteria pollutants,
additional vehicle use associated with the rebound effect will increase
emissions of these pollutants. Thus the net effect of stricter
standards on emissions of each criteria pollutant depends on the
relative magnitudes of reduced emissions from fuel refining and
distribution, and increases in emissions resulting from added vehicle
use. Criteria air pollutants emitted by vehicles and during fuel
production include carbon monoxide (CO), hydrocarbon compounds (usually
referred to as ``volatile organic compounds,'' or VOC), nitrogen oxides
(NOX), fine particulate matter (PM2.5), and
sulfur oxides (SOX). It is assumed that the emission rates
(per mile) stay constant for future year vehicles.
[cir] Economic value of reductions in criteria air pollutants--For
the purpose of the joint technical analysis, EPA and NHTSA estimate the
economic value of the human health benefits associated with reducing
exposure to PM2.5 using a ``benefit-per-ton'' method. These
PM2.5-related benefit-per-ton estimates provide the total
monetized benefits to human health (the sum of reductions in premature
mortality and premature morbidity) that result from eliminating one ton
of directly emitted PM2.5, or one ton of a pollutant that
contributes to secondarily-formed PM2.5 (such as
NOX, SOX, and VOCs), from a specified source.
Chapter 4.2.9 of the Technical Support Document that accompanies this
rule includes a description of these values. Separately, EPA also
conducted air quality modeling to estimate the change in ambient
concentrations of criteria pollutants and used this as a basis for
estimating the human health benefits and their economic value. Section
III.H.7 presents these benefits estimates.
[cir] Reductions in GHG emissions--Emissions of carbon dioxide and
other GHGs occur throughout the process of producing and distributing
transportation fuels, as well as from fuel combustion itself. By
reducing the volume of fuel consumed by passenger cars and light
trucks, higher standards will thus reduce GHG emissions generated by
fuel use, as well as throughout the fuel supply cycle. The agencies
estimated the increases of GHGs other than CO2, including
methane and nitrous oxide, from additional vehicle use by multiplying
the increase in total miles driven by cars and light trucks of each
model year and age by emission rates per vehicle-mile for these GHGs.
These emission rates, which differ between cars and light trucks as
well as between gasoline and diesel vehicles, were estimated by EPA
using its recently-developed Motor Vehicle Emission Simulator (Draft
MOVES 2010).\108\ Increases in emissions of non-CO2 GHGs are
converted to equivalent increases in CO2 emissions using
estimates of the Global Warming Potential (GWP) of methane and nitrous
oxide.
---------------------------------------------------------------------------

\108\ The MOVES model assumes that the per-mile rates at which
cars and light trucks emit these GHGs are determined by the
efficiency of fuel combustion during engine operation and chemical
reactions that occur during catalytic after-treatment of engine
exhaust, and are thus independent of vehicles' fuel consumption
rates. Thus MOVES' emission factors for these GHGs, which are
expressed per mile of vehicle travel, are assumed to be unaffected
by changes in fuel economy.
---------------------------------------------------------------------------

[cir] Economic value of reductions in CO2 emissions --
EPA and NHTSA assigned a dollar value to reductions in CO2
emissions using the marginal dollar value (i.e., cost) of climate-
related damages resulting from carbon emissions, also referred to as
``social cost of carbon'' (SCC). The SCC is intended to measure the
monetary value society places on impacts resulting from increased GHGs,
such as property damage from sea level rise, forced migration due to
dry land loss, and mortality changes associated with vector-borne
diseases. Published estimates of the SCC vary widely as a result of
uncertainties about future economic growth, climate sensitivity to GHG
emissions, procedures used to model the economic impacts of climate
change, and the choice of discount rates.
EPA and NHTSA received extensive comments about how to improve the
characterization of the SCC and have since developed new estimates
through an interagency modeling exercise. The comments addressed
various issues, such as discount rate selection, treatment of
uncertainty, and emissions and socioeconomic trajectories, and
justified the revision of SCC for the final rule. The modeling exercise
involved running three integrated assessment models using inputs agreed
upon by the interagency group for climate sensitivity, socioeconomic
and emissions trajectories, and discount rates. A more complete
discussion of SCC can be found in the Technical Support Document,
Social Cost of Carbon for Regulatory Impact Analysis Under Executive
Order 12866 (hereafter, ``SCC TSD''); revised SCC estimates
corresponding to assumed values of the discount rate are shown in Table
II.F-1.\109\
---------------------------------------------------------------------------

\109\ Interagency Working Group on Social Cost of Carbon, U.S.
Government, with participation by Council of Economic Advisers,
Council on Environmental Quality, Department of Agriculture,
Department of Commerce, Department of Energy, Department of
Transportation, Environmental Protection Agency, National Economic
Council, Office of Energy and Climate Change, Office of Management
and Budget, Office of Science and Technology Policy, and Department
of Treasury, ``Social Cost of Carbon for Regulatory Impact Analysis
Under Executive Order 12866,'' February 2010, available in docket
EPA-HQ-OAR-2009-0472.

[[Page 25381]]

Table II.F-1--Social Cost of CO2, 2010
[In 2007 dollars]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Discount Rate 5% 3% 2.5% 3%
--------------------------------------------------------------------------------------------------------------------------------------------------------
Source of Estimate............................ Mean of Estimates Values 95th percentile estimate.
--------------------------------------------------------------------------------------------------------------------------------------------------------
2010 Estimate................................. $5 $21 $35 $65.
--------------------------------------------------------------------------------------------------------------------------------------------------------

Discounting future benefits and costs--Discounting future
fuel savings and other benefits is intended to account for the
reduction in their value to society when they are deferred until some
future date, rather than received immediately. The discount rate
expresses the percent decline in the value of these benefits--as viewed
from today's perspective--for each year they are deferred into the
future. In evaluating the non-climate related benefits of the final
standards, the agencies have employed discount rates of both 3 percent
and 7 percent. We received some comments on the discount rates used in
the proposal, most of which were directed at the discount rates used to
value future fuel savings and the rates used to value of the social
cost of carbon. In general, commenters were supporting one of the
discount rates over the other, although some suggested that our rates
were too high or too low. We have revised the discounting used when
calculating the net present value of social cost of carbon as explained
in Sections III.H. and VI but have not revised our discounting
procedures for other costs or benefits.
For the reader's reference, Table II.F-2 below summarizes the
values used to calculate the impacts of each final standard. The values
presented in this table are summaries of the inputs used for the
models; specific values used in the agencies' respective analyses may
be aggregated, expanded, or have other relevant adjustments. See the
respective RIAs for details.
The agencies recognize that each of these values has some degree of
uncertainty, which the agencies further discuss in the Joint TSD. The
agencies have conducted a range of sensitivities and present them in
their respective RIAs. For example, NHTSA has conducted a sensitivity
analysis on several assumptions including (1) forecasts of future fuel
prices, (2) the discount rate applied to future benefits and costs, (3)
the magnitude of the rebound effect, (4) the value to the U.S. economy
of reducing carbon dioxide emissions, (5) inclusion of the monopsony
effect, and (6) the reduction in external economic costs resulting from
lower U.S. oil imports. This information is provided in NHTSA's RIA.

Table II.F-2--Economic Values for Benefits Computations
[2007$]
------------------------------------------------------------------------

------------------------------------------------------------------------
Fuel Economy Rebound Effect............... 10%.
``Gap'' between test and on-road MPG...... 20%.
Value of refueling time per ($ per vehicle- $24.64.
hour).
Average tank volume refilled during 55%.
refueling stop.
Annual growth in average vehicle use...... 1.15%.
Fuel Prices (2012-50 average, $/gallon): ............................
Retail gasoline price................. $3.66.
Pre-tax gasoline price................ $3.29.
------------------------------------------------------------------------
Economic Benefits From Reducing Oil Imports ($/gallon)
------------------------------------------------------------------------
``Monopsony'' Component................... $0.00.
Price Shock Component..................... $0.17.
Military Security Component............... $0.00.
Total Economic Costs ($/gallon)........... $0.17.
------------------------------------------------------------------------
Emission Damage Costs (2020, $/ton or $/metric ton)
------------------------------------------------------------------------
Carbon monoxide........................... $0.
Volatile organic compounds (VOC).......... $1,300.
Nitrogen oxides (NOX)--vehicle use........ $5,100.
Nitrogen oxides (NOX)--fuel production and $ 5,300.
distribution.
Particulate matter (PM2.5)--vehicle use... $ 240,000.
Particulate matter (PM2.5)--fuel $ 290,000.
production and distribution.
Sulfur dioxide (SO2)...................... $ 31,000.
Carbon dioxide (CO2) emissions in 2010.... $5.
$21.
$35.
$65.
Annual Increase in CO2 Damage Cost........ variable, depending on
estimate.
------------------------------------------------------------------------
External Costs From Additional Automobile Use ($/vehicle-mile)
------------------------------------------------------------------------
Congestion................................ $ 0.054.
Accidents................................. $ 0.023.
Noise..................................... $ 0.001.
-----------------------------

[[Page 25382]]

Total External Costs.................. $ 0.078.
------------------------------------------------------------------------
External Costs From Additional Light Truck Use ($/vehicle-mile)
------------------------------------------------------------------------
Congestion................................ $0.048.
Accidents................................. $0.026.
Noise..................................... $0.001.
Total External Costs...................... $0.075.
Discount Rates Applied to Future Benefits. 3%, 7%.
------------------------------------------------------------------------

G. What are the estimated safety effects of the final MYs 2012-2016
CAFE and GHG standards?

The primary goals of the final CAFE and GHG standards are to reduce
fuel consumption and GHG emissions, but in addition to these intended
effects, the agencies must consider the potential of the standards to
affect vehicle safety,\110\ which the agencies have assessed in
evaluating the appropriate levels at which to set the final standards.
Safety trade-offs associated with fuel economy increases have occurred
in the past, and the agencies must be mindful of the possibility of
future ones. These past safety trade-offs occurred because
manufacturers chose, at the time, to build smaller and lighter
vehicles--partly in response to CAFE standards--rather than adding more
expensive fuel-saving technologies (and maintaining vehicle size and
safety), and the smaller and lighter vehicles did not fare as well in
crashes as larger and heavier vehicles. Historically, as shown in FARS
data analyzed by NHTSA, the safest vehicles have been heavy and large,
while the vehicles with the highest fatal-crash rates have been light
and small, both because the crash rate is higher for small/light
vehicles and because the fatality rate per crash is higher for small/
light vehicle crashes.
---------------------------------------------------------------------------

\110\ In this rulemaking document, vehicle safety is defined as
societal fatality rates which include fatalities to occupants of all
the vehicles involved in the collisions, plus any pedestrians.
---------------------------------------------------------------------------

Changes in relative safety are related to shifts in the
distribution of vehicles on the road. A policy that induces a widening
in the size distribution of vehicles on the road, could result in
negative impacts on safety, The primary mechanism in this rulemaking
for mitigating the potential negative effects on safety is the
application of footprint-based standards, which create a disincentive
for manufacturers to produce smaller-footprint vehicles. This is
because as footprint decreases, the corresponding fuel economy/GHG
emission target becomes more stringent.\111\ The shape of the footprint
curves themselves have also been designed to be approximately
``footprint neutral'' within the sloped portion of the functions--that
is, to neither encourage manufacturers to increase the footprint of
their fleets, nor to decrease it. Upsizing also is discouraged through
a ``cut-off'' at larger footprints. For both cars and light trucks
there is a ``cut-off'' that affects vehicles smaller than 41 square
feet. The agencies recognize that for manufacturers who make small
vehicles in this size range, this cut off creates some incentive to
downsize (i.e. further reduce the size and/or increase the production
of models currently smaller than 41 square feet) to make it easier to
meet the target. The cut off may also create some incentive for
manufacturers who do not currently offer such models to do so in the
future. However, at the same time, the agencies believe that there is a
limit to the market for cars smaller than 41 square feet--most
consumers likely have some minimum expectation about interior volume,
among other things. In addition, vehicles in this market segment are
the lowest price point for the light-duty automotive market, with a
number of models in the $10,000 to $15,000 range. In order to justify
selling more vehicles in this market in order to generate fuel economy
or CO2 credits (that is, for this final rule to be the
incentive for selling more vehicles in this small car segment), a
manufacturer would need to add additional technology to the lowest
price segment vehicles, which could be challenging. Therefore, due to
these two reasons (a likely limit in the market place for the smallest
sized cars and the potential consumer acceptance difficulty in adding
the necessary technologies in order to generate fuel economy and
CO2 credits), the agencies believe that the incentive for
manufacturers to increase the sale of vehicles smaller than 41 square
feet due to this rulemaking, if present, is small. For further
discussion on these aspects of the standards, please see Section II.C
above and Chapter 2 of the Joint TSD.
---------------------------------------------------------------------------

\111\ We note, however, that vehicle footprint is not synonymous
with vehicle size. Since the footprint is only that portion of the
vehicle between the front and rear axles, footprint-based standards
do not discourage downsizing the portions of a vehicle in front of
the front axle and to the rear of the rear axle, or to other
portions of the vehicle outside the wheels. The crush space provided
by those portions of a vehicle can make important contributions to
managing crash energy. At least one manufacturer has confidentially
indicated plans to reduce overhang as a way of reducing mass on some
vehicles during the rulemaking time frame. Additionally, simply
because footprint-based standards create no incentive to downsize
vehicles, does not mean that manufacturers may not choose to do so
if doing so makes it easier to meet the overall standard (as, for
example, if the smaller vehicles are so much lighter that they
exceed their targets by much greater amounts).
---------------------------------------------------------------------------

Manufacturers have stated, however, that they will reduce vehicle
weight as one of the cost-effective means of increasing fuel economy
and reducing CO2 emissions, and the agencies have
incorporated this expectation into our modeling analysis supporting
today's final standards. NHTSA's previous analyses examining the
relationship between vehicle mass and fatalities found fatality
increases as vehicle weight and size were reduced, but these previous
analyses did not differentiate between weight reductions and size
(i.e., weight and footprint) reductions.
The question of the effect of changes in vehicle mass on safety in
the context of fuel economy is a complex question that poses serious
analytic challenges and has been a contentious issue for many years, as
discussed by a number of commenters to the NPRM. This contentiousness
arises, at least in part, from the difficulty of isolating vehicle mass
from other confounding factors (e.g., driver behavior, or vehicle
factors such as engine size and wheelbase). In addition, several
vehicle factors have been closely related historically, such as vehicle
mass, wheelbase, and track width. The issue has been reviewed and
analyzed in the literature for more than two decades. For the reader's
reference, much more information about safety in the CAFE context is
available in Chapter IX of NHTSA's FRIA. Chapter 7.6 of EPA's final RIA
also contained

[[Page 25383]]

additional discussion on mass and safety.
Over the past several years, as also discussed by a number of
commenters to the NPRM, contention has arisen with regard to the
applicability of analysis of historical crash data to future safety
effects due to mass reduction. The agencies recognize that there are a
host of factors that may make future mass reduction different than what
is reflected in the historical data. For one, the footprint-based
standards have been carefully developed by the agencies so that they do
not encourage vehicle footprint reductions as a way of meeting the
standards, but so that they do encourage application of fuel-saving
technologies, including mass reduction. This in turn encourages
manufacturers to find ways to separate mass reduction from footprint
reduction, which will very likely result in a future relationship
between mass and fatalities that is safer than the historical
relationship. However, as manufacturers pursue these methods of mass
reduction, the fleet moves further away from the historical trends,
which the agencies recognize.
NHTSA's NPRM analysis of the safety effects of the proposed CAFE
standards was based on NHTSA's 2003 report concerning mass and size
reduction in MYs 1991-1999 vehicles, and evaluated a ``worst-case
scenario'' in which the safety effects of the combined reductions of
both mass and size for those vehicles were determined for the future
passenger car and light truck fleets.\112\ In the NPRM analysis, mass
and size could not be separated from one another, resulting in what
NHTSA recognized was a larger safety disbenefit than was likely under
the MYs 2012-2016 footprint-based CAFE standards. NHTSA emphasized,
however, that actual fatalities would likely be less than these
``worst-case'' estimates, and possibly significantly less, based on the
various factors discussed in the NPRM that could reduce the estimates,
such as careful mass reduction through material substitution, etc.
---------------------------------------------------------------------------

\112\ The analysis excluded 2-door cars.
---------------------------------------------------------------------------

For the final rule, as discussed in the NPRM and in recognition of
the importance of conducting analysis that better reflects, within the
limits of our current knowledge, the potential safety effects of future
mass reduction in response to the final CAFE and GHG standards that is
highly unlikely to involve concurrent reductions in footprint, NHTSA
has revised its analysis in consultation with EPA. Perhaps the most
important change has been that NHTSA agreed with commenters that it was
both possible and appropriate to separate the effect of mass reductions
from the effect of footprint reductions. NHTSA thus performed a new
statistical analysis, hereafter referred to as the 2010 Kahane
analysis, of the MYs 1991-99 vehicle database from its 2003 report (now
including rather than excluding 2-door cars in the passenger car
fleet), assessing relationships between fatality risk, mass, and
footprint for both passenger cars and LTVs (light trucks and
vans).\113\ As part of its results, the new report presents an ``upper-
estimate scenario,'' a ``lower-estimate scenario,'' as well as an
``actual regression result scenario'' representing potential safety
effects of future mass reductions without corresponding vehicle size
reductions, that assume, by virtue of being a cross-sectional analysis
of historical data, that historical relationships between vehicle mass
and fatalities are maintained. The ``upper-estimate scenario'' and
``lower-estimate scenario'' are based on NHTSA's judgment as a vehicle
safety agency, and are not meant to convey any more or less likelihood
in the results, but more to convey a sense of bounding for potential
safety effects of reducing mass while holding footprint constant. The
upper-estimate scenario reflects potential safety effects given the
report's finding that, using the one-step regression method of the 2003
Kahane report, the regression coefficients show that mass and footprint
each accounted for about half the fatality increase associated with
downsizing in a cross-sectional analysis of MYs 1991-1999 cars. A
similar effect was found for lighter LTVs. Applying the same regression
method to heavier LTVs, however, the coefficients indicated a
significant societal fatality reduction when mass, but not footprint,
is reduced in the heavier LTVs.\114\ Fatalities are reduced primarily
because mass reduction in the heavier LTVs will reduce risk to
occupants of the other cars and lighter LTVs involved in collisions
with these heavier LTVs.\115\ Thus, even in the ``upper-estimate
scenario,'' the potential fatality increases associated with mass
reduction in the passenger cars would be to a large extent offset by
the benefits of mass reduction in the heavier LTVs.
---------------------------------------------------------------------------

\113\ ``Relationships Between Fatality Risk, Mass, and Footprint
in Model Year 1991-1999 and Other Passenger Cars and LTVs,'' Charles
J. Kahane, NCSA, NHTSA, March 2010. The text of the report may be
found in Chapter IX of NHTSA's FRIA, where it constitutes a section
of that chapter. We note that this report has not yet been
externally peer-reviewed, and therefore may be changed or refined
after it has been subjected to peer review. The results of the
report have not been included in the tables summarizing the costs
and benefits of this rulemaking and did not affect the stringency of
the standards. NHTSA has begun the process for obtaining peer review
in accordance with OMB guidance. The agency will ensure that
concerns raised during the peer review process are addressed before
relying on the report for future rulemakings. The results of the
peer review and any subsequent revisions to the report will be made
available in a public docket and on NHTSA's Web site as they are
completed.
\114\ Conversely, the coefficients indicate a significant
increase if footprint is reduced.
\115\ We note that there may be some (currently non-
quantifiable) welfare losses for purchasers of these heavier LTVs,
the mass of which is reduced in response to these final standards.
This is due to the fact that in certain crashes, as discussed below
and in greater detail in Chapter IX of the NHTSA FRIA, more mass
will always be helpful (although certainly in other crashes, the
amount of mass reduction modeled by the agency will not be enough to
have any significant effect on driver/occupant safety). However, we
believe the effects of this will likely be minor. Consumer welfare
impacts of the final rule are discussed in more detail in Chapter
VIII of the NHTSA FRIA.
---------------------------------------------------------------------------

The lower-estimate scenario, in turn, reflects NHTSA's estimate of
potential safety effects if future mass reduction is accomplished
entirely by material substitution, smart design,\116\ and component
integration, among other things, that can reduce mass without
perceptibly changing a vehicle's shape, functionality, or safety
performance, maintaining structural strength without compromising other
aspects of safety. If future mass reduction follows this path, it could
limit the added risk close to only the effects of mass per se (the
ability to transfer momentum to other vehicles or objects in a
collision), resulting in estimated effects in passenger cars that are
substantially smaller than in the upper-estimate scenario based
directly on the regression results. The lower-estimate scenario also
covers both passenger cars and LTVs.
---------------------------------------------------------------------------

\116\ Manufacturers may reduce mass through smart design using
computer aided engineering (CAE) tools that can be used to better
optimize load paths within structures by reducing stresses and
bending moments applied to structures. This allows better
optimization of the sectional thicknesses of structural components
to reduce mass while maintaining or improving the function of the
component. Smart designs also integrate separate parts in a manner
that reduces mass by combining functions or the reduced use of
separate fasteners. In addition, some ``body on frame'' vehicles are
redesigned with a lighter ``unibody'' construction.
---------------------------------------------------------------------------

Overall, based on the new analyses, NHTSA estimated that fatality
effects could be markedly less than those estimated in the ``worst-case
scenario'' presented in the NPRM. The agencies believe that the overall
effect of mass reduction in cars and LTVs may be close to zero, and may
possibly be beneficial in terms of the fleet as a whole if mass
reduction is carefully applied in the future (as with careful material
substitution and other methods of mass reduction that can reduce mass
without perceptibly changing a car's shape, functionality, or safety
performance,

[[Page 25384]]

and maintain its structural strength without making it excessively
rigid). This is especially important if the mass reduction in the
heavier LTVs is greater (in absolute terms) than in passenger cars, as
discussed further below and in the 2010 Kahane report.
The following sections will address how the agencies addressed
potential safety effects in the NPRM for the proposed standards, how
commenters responded, and the work that NHTSA has done since the NPRM
to revise its estimates of potential safety effects for the final rule.
The final section discusses some of the agencies' plans for the future
with respect to potential analysis and studies to further enhance our
understanding of this important and complex issue.
1. What did the agencies say in the NPRM with regard to potential
safety effects?
In the NPRM preceding these final standards, NHTSA's safety
assessment derived from the agency's belief that some of these vehicle
factors, namely vehicle mass and footprint, could not be accurately
separated. NHTSA relied on the 2003 study by Dr. Charles Kahane, which
estimates the effect of 100-pound reductions in MYs 1991-1999 heavy
light trucks and vans (LTVs), light LTVs, heavy passenger cars, and
light passenger cars.\117\ The study compares the fatality rates of
LTVs and cars to quantify differences between vehicle types, given
drivers of the same age/gender, etc. In that analysis, the effect of
``weight reduction'' is not limited to the effect of mass per se, but
includes all the factors, such as length, width, structural strength,
safety features, and size of the occupant compartment, that were
naturally or historically confounded with mass in MYs 1991-1999
vehicles. The rationale was that adding length, width, or strength to a
vehicle historically also made it heavier.
---------------------------------------------------------------------------

\117\ Kahane, Charles J., PhD, ``Vehicle Weight, Fatality Risk
and Crash Compatibility of Model Year 1991-99 Passenger Cars and
Light Trucks,'' DOT HS 809 662, October 2003, Executive Summary.
Available at http://www.nhtsa.dot.gov/cars/rules/regrev/evaluate/
809662.html (last accessed March 10, 2010).
---------------------------------------------------------------------------

NHTSA utilized the relationships between mass and safety from
Kahane (2003), expressed as percentage increases in fatalities per 100-
pound mass reduction, and examined the mass effects assumed in the NPRM
modeling analysis. While previous CAFE rulemakings had limited mass
reduction as a ``technology option'' to vehicles over 5,000 pounds
GVWR, both NHTSA's and EPA's modeling analyses in the NPRM included
mass reduction of up to 5-10 percent of baseline curb weight, depending
on vehicle subclass, in response to recently-submitted manufacturer
product plans as well as public statements indicating that these levels
were possible and likely. 5-10 percent represented a maximum bound;
EPA's modeling, for example, included average vehicle weight reductions
of 4 percent between MYs 2011 and 2016, although the average per-
vehicle mass reduction was greater in absolute terms for light trucks
than for passenger cars. NHTSA's assumptions for mass reduction were
also limited by lead time such that mass reductions of 1.5 percent were
included for redesigns occurring prior to MY 2014, and mass reductions
of 5-10 percent were only ``achievable'' in redesigns occurring in MY
2014 or later. NHTSA further assumed that mass reductions would be
limited to 5 percent for small vehicles (e.g., subcompact passenger
cars), and that reductions of 10 percent would only be applied to the
larger vehicle types (e.g., large light trucks).
Based on these assumptions of how manufacturers might comply with
the standards, NHTSA examined the effects of the identifiable safety
trends over the lifetime of the vehicles produced in each model year.
The effects were estimated on a year-by-year basis, assuming that
certain known safety trends would result in a reduction in the target
population of fatalities from which the mass effects are derived.\118\
Using this method, NHTSA found a 12.6 percent reduction in fatality
levels between 2007 and 2020. The estimates derived from applying
Kahane's 2003 percentages to a baseline of 2007 fatalities were then
multiplied by 0.874 to account for changes that the agency believed
would take place in passenger car and light truck safety between the
2007 baseline on-road fleet used for that particular analysis and year
2020.\119\
---------------------------------------------------------------------------

\118\ NHTSA explained that there are several identifiable safety
trends that are already in place or expected to occur in the
foreseeable future and that were not accounted for in the study. For
example, two important new safety standards that have already been
issued and will be phasing in during the rulemaking time frame.
Federal Motor Vehicle Safety Standard No. 126 (49 CFR 571.126) will
require electronic stability control in all new vehicles by MY 2012,
and the upgrade to Federal Motor Vehicle Safety Standard No. 214
(Side Impact Protection, 49 CFR 571.214) will likely result in all
new vehicles being equipped with head-curtain air bags by MY 2014.
Additionally, the agency stated that it anticipates continued
improvements in driver (and passenger) behavior, such as higher
safety belt use rates. All of these will tend to reduce the absolute
number of fatalities resulting from mass reductions. Thus, while the
percentage increases in Kahane (2003) was applied, the reduced base
resulted in smaller absolute increases than those that were
predicted in the 2003 report.
\119\ Blincoe, L. and Shankar, U, ``The Impact of Safety
Standards and Behavioral Trends on Motor Vehicle Fatality Rates,''
DOT HS 810 777, January 2007. See Table 4 comparing 2020 to 2007
(37,906/43,363 = 12.6% reduction (1-.126 = .874)
---------------------------------------------------------------------------

NHTSA and EPA both emphasized that the safety effect estimates in
the NPRM needed to be understood in the context of the 2003 Kahane
report, which is based upon a cross-sectional analysis of the actual
on-road safety experience of 1991-1999 vehicles. For those vehicles,
heavier usually also meant larger-footprint. Hence, the numbers in
those analyses were used to predict the safety-related fatalities that
could occur in the unlikely event that weight reduction for MYs 2012-
2016 is accomplished entirely by reducing mass and reducing footprint.
Any estimates derived from those analyses represented a ``worst-case''
estimate of safety effects, for several reasons.
First, manufacturers are far less likely to reduce mass by
``downsizing'' (making vehicles smaller overall) under the current
attribute-based standards, because the standards are based on vehicle
footprint. The selection of footprint as the attribute in setting CAFE
and GHG standards helps to reduce the incentive to alter a vehicle's
physical dimensions. This is because as footprint decreases, the
corresponding fuel economy/GHG emission target becomes more
stringent.\120\ The shape of the footprint curves themselves have also
been designed to be approximately ``footprint neutral'' within the
sloped portion of the functions--that is, to neither encourage
manufacturers to increase the footprint of their fleets, nor to
decrease it. For further discussion on these aspects of the standards,
please see Section II.C above and Chapter 2 of the Joint TSD. However,
as discussed in Sections III.H.1 and IV.G.6 below, the agencies
acknowledge some uncertainty regarding how consumer purchases will
change in response to the vehicles

[[Page 25385]]

designed to meet the MYs 2012-2016 standards. This could potentially
affect the mix of vehicles sold in the future, including the mass and
footprint distribution.
---------------------------------------------------------------------------

\120\ We note, however, that vehicle footprint is not synonymous
with vehicle size. Since the footprint is only that portion of the
vehicle between the front and rear axles, footprint-based standards
do not discourage downsizing the portions of a vehicle in front of
the front axle and to the rear of the rear axle, or to other
portions of the vehicle outside the wheels. The crush space provided
by those portions of a vehicle can make important contributions to
managing crash energy. NHTSA noted in the NPRM that at least one
manufacturer has confidentially indicated plans to reduce overhang
as a way of reducing mass on some vehicles during the rulemaking
time frame. Additionally, simply because footprint-based standards
create no incentive to downsize vehicles, does not mean that
manufacturers may not choose to do so if doing so makes it easier to
meet the overall standard (as, for example, if the smaller vehicles
are so much lighter that they exceed their targets by much greater
amounts).
---------------------------------------------------------------------------

As a result, the agencies found it likely that a significant
portion of the mass reduction in the MY 2012-2016 vehicles would be
accomplished by strategies, such as material substitution, smart
design, reduced powertrain requirements,\121\ and mass compounding,
that have a lesser safety effect than the prevalent 1980s strategy of
simply making the vehicles smaller. The agencies noted that to the
extent that future mass reductions could be achieved by these methods--
without any accompanying reduction in the size or structural strength
of the vehicle--then the fatality increases associated with the mass
reductions anticipated by the model as a result of the proposed
standards could be significantly smaller than those in the worst-case
scenario.
---------------------------------------------------------------------------

\121\ Reduced powertrain requirements do not include a reduction
in performance. When vehicle mass is reduced, engine torque and
transmission gearing can be altered so that acceleration performance
is held constant instead of improving. A detailed discussion is
included in Chapter 3 of the Technical Support Document.
---------------------------------------------------------------------------

However, even though the agencies recognized that these methods of
mass reduction could be technologically feasible in the rulemaking time
frame, and included them as such in our modeling analyses, the agencies
diverged as to how potential safety effects accompanying such methods
of mass reduction could be evaluated, particularly in relation to the
worst-case scenario presented by NHTSA. NHTSA stated that it could not
predict how much smaller those increases would be for any given mixture
of mass reduction methods, since the data on the safety effects of mass
reduction alone (without size reduction) was not available due to the
low numbers of vehicles in the current on-road fleet that have utilized
these technologies extensively. Further, to the extent that mass
reductions were accomplished through use of light, high-strength
materials, NHTSA emphasized that there would be significant additional
costs that would need to be determined and accounted for than were
reflected in the agency's proposal.
Additionally, NHTSA emphasized that while it thought material
substitution and other methods of mass reduction could considerably
lessen the potential safety effects compared to the historical trend,
NHTSA also stated that it did not believe the effects in passenger cars
would be smaller than zero. EPA disagreed with this, and stated in the
NPRM that the safety effects could very well be smaller than zero. Even
though footprint-based standards discourage downsizing as a way of
``balancing out'' sales of larger/heavier vehicles, they do not
discourage manufacturers from reducing crush space in overhang areas or
from reducing structural support as a way of taking out mass.\122\
Moreover, NHTSA's analysis had also found that lighter cars have a
higher involvement rate in fatal crashes, even after controlling for
the driver's age, gender, urbanization, and region of the country.
Being unable to explain this clear trend in the crash data, NHTSA
stated that it must assume that mass reduction is likely to be
associated with higher fatal-crash rates, no matter how the weight
reduction is achieved.
---------------------------------------------------------------------------

\122\ However, we recognize that FMVSS and NCAP ratings may
limit the manufacturer's ability to reduce crush space or structural
support.
---------------------------------------------------------------------------

NHTSA also noted in the NPRM that several studies by Dynamic
Research, Inc. (DRI) had been repeatedly cited to the agency in support
of the proposition that reducing vehicle mass while maintaining track
width and wheelbase would lead to significant safety benefits. In its
2005 studies, one of which was published and peer-reviewed through the
Society of Automotive Engineers as a technical paper, DRI attempted to
assess the independent effects of vehicle weight and size (in terms of
wheelbase and track width) on safety, and presented results indicating
that reducing vehicle weight tends to reduce fatalities, but that
reducing vehicle wheelbase and track width tends to increase
fatalities. DRI's analysis was based on FARS data for MYs 1985-1998
passenger cars and 1985-1997 light trucks, similar to the MYs 1991-1999
car and truck data used in the 2003 Kahane report. However, DRI
included 2-door passenger cars, while the 2003 Kahane report excluded
those vehicles out of concern that their inclusion could bias the
results of the regression analysis, because a significant proportion of
MYs 1991-1999 2-door cars were sports and ``muscle'' cars, which have
particularly high fatal crash rates for their relatively short
wheelbases compared to the rest of the fleet. While in the NPRM NHTSA
rejected the results of the DRI studies based in part on this concern,
the agencies note that upon further consideration, NHTSA has agreed for
this final rule that the inclusion of 2-door cars in regression
analysis of historical data is appropriate, and indeed has no overly-
biasing effects.
The 2005 DRI studies also differed from the 2003 Kahane report in
terms of their estimates of the effect of vehicle weight on rollover
fatalities. The 2003 Kahane report analyzed a single variable, curb
weight, as a surrogate for both vehicle size and weight, and found that
curb weight reductions would increase rollover fatalities. The DRI
study, in contrast, attempted to analyze curb weight, wheelbase, and
track width separately, and found that curb weight reduction would
decrease rollover fatalities, while wheelbase reduction and track width
reduction would increase them. DRI suggested that heavier vehicles may
have higher rollover fatalities for two reasons: first, because taller
vehicles tend to be heavier, so the correlation between vehicle height
and weight and vehicle center-of-gravity height may make heavier
vehicles more rollover-prone; and second, because heavier vehicles may
have been less rollover-crashworthy due to FMVSS No. 216's constant (as
opposed to proportional) requirements for MYs 1995-1999 vehicles
weighing more than 3,333 lbs unloaded.
Overall, DRI's 2005 studies found a reduction in fatalities for
cars (580 in the first study, and 836 in the second study) and for
trucks (219 in the first study, 682 in the second study) for a 100
pound reduction in curb weight without accompanying wheelbase or track
width reductions. In the NPRM, NHTSA disagreed with the results of the
DRI studies, out of concern that DRI's inclusion of 2-door cars in its
analysis biased the results, and because NHTSA was unable to reproduce
DRI's results despite repeated attempts. NHTSA stated that it agreed
intuitively with DRI's conclusion that vehicle mass reductions without
accompanying size reductions (as through substitution of a heavier
material for a lighter one) would be less harmful than downsizing, but
without supporting real-world data and unable to verify DRI's results,
NHTSA stated that it could not conclude that mass reductions would
result in safety benefits. EPA, in contrast, believed that DRI's
results contained some merit, in particular because the study separated
the effects of mass and size and EPA stated that applying them using
the curb weight reductions in EPA's modeling analysis would show an
overall reduction of fatalities for the proposed standards.
On balance, both agencies recognized that mass reduction could be
an important tool for achieving higher levels of fuel economy and
reducing CO2 emissions, and emphasized that NHTSA's fatality
estimates represented a worst-case scenario for the potential effects
of the proposed standards, and

[[Page 25386]]

that actual fatalities will be less than these estimates, possibly
significantly less, based on the various factors discussed in the NPRM
that could reduce the estimates. The agencies sought comment on the
safety analysis and discussions presented in the NPRM.
2. What public comments did the agencies receive on the safety analysis
and discussions in the NPRM?
Several dozen commenters addressed the safety issue. Claims and
arguments made by commenters in response to the safety effects analysis
and discussion in the NPRM tended to follow several general themes, as
follows:
NHTSA's safety effects estimates are inaccurate because
they do not account for:
[cir] While NHTSA's study only considers vehicles from MYs 1991-
1999, more recently-built vehicles are safer than those, and future
vehicles will be safer still;
[cir] Lighter vehicles are safer than heavier cars in terms of
crash-avoidance, because they handle and brake better;
[cir] Fatalities are linked more to other factors than mass;
[cir] The structure of the standards reduces/contributes to
potential safety effects from mass reduction;
[cir] NHTSA could mitigate additional safety effects from mass
reduction, if there are any, by simply regulating safety more;
[cir] Casualty risks range widely for vehicles of the same weight
or footprint, which skews regression analysis and makes computer
simulation a better predictor of the safety effects of mass reduction;
DRI's analysis shows that lighter vehicles will save
lives, and NHTSA reaches the opposite conclusion without disproving
DRI's analysis;
[cir] Possible reasons that NHTSA and DRI have reached different
conclusions:
[dec222] NHTSA's study should distinguish between reductions in
size and reductions in weight like DRI's;
[dec222] NHTSA's study should include two-door cars;
[dec222] NHTSA's study should have used different assumptions;
[dec222] NHTSA's study should include confidence intervals;
NHTSA should include a ``best-case'' estimate in its
study;
NHTSA should not include a ``worst-case'' estimate in its
study;
The agencies recognize that the issue of the potential safety
effects of mass reduction, which was one of the many factors considered
in the balancing that led to the agencies' conclusion as to appropriate
stringency levels for the MYs 2012-2016 standards, is of great interest
to the public and could possibly be a more significant factor in
regulators' and manufacturers' decisions with regard to future
standards beyond MY 2016. The agencies are committed to analyzing this
issue thoroughly and holistically going forward, based on the best
available science, in order to further their closely related missions
of safety, energy conservation, and environmental protection. We
respond to the issues and claims raised by commenters in turn below.

NHTSA's estimates are inaccurate because NHTSA's study only considers
vehicles from MYs 1991-1999, but more recently-built vehicles are safer
than those, and future vehicles will be safer still

A number of commenters (CAS, Adcock, NACAA, NJ DEP, NY DEC, UCS,
and Wenzel) argued that the 2003 Kahane report, on which the ``worst-
case scenario'' in the NPRM was based, is outdated because it considers
the relationship between vehicle weight and safety in MYs 1991-1999
passenger cars. These commenters generally stated that data from MYs
1991-1999 vehicles provide an inaccurate basis for assessing the
relationship between vehicle weight and safety in current or future
vehicles, because the fleets of vehicles now and in the future are
increasingly different from that 1990s fleet (more crossovers, fewer
trucks, lighter trucks, etc.), with different vehicle shapes and
characteristics, different materials, and more safety features. Several
of these commenters argued that NHTSA should conduct an updated
analysis for the final rule using more recent data--Wenzel, for
example, stated that an updated regression analysis that accounted for
the recent introduction of crossover SUVs would likely find reduced
casualty risk, similar to DRI's previous finding using fatality data.
CEI, in contrast, argued that the ``safety trade-off'' would not be
eliminated by new technologies and attribute-based standards, because
additional weight inherently makes a vehicle safer to its own
occupants, citing the 2003 Kahane report, while AISI argued that
Desapriya had found that passenger car drivers and occupants are two
times more likely to be injured than drivers and occupants in larger
pickup trucks and SUVs.
Several commenters (Adcock, CARB, Daimler, NESCAUM, NRDC, Public
Citizen, UCS, Wenzel) suggested that NHTSA's analysis was based on
overly pessimistic assumptions about how manufacturers would choose to
reduce mass in their vehicles, because manufacturers have a strong
incentive in the market to build vehicles safely. Many of these
commenters stated that several manufacturers have already committed
publicly to fairly ambitious mass reduction goals in the mid-term, but
several stated further that NHTSA should not assume that manufacturers
will reduce the same amount of mass in all vehicles, because it is
likely that they will concentrate mass reduction in the heaviest
vehicles, which will improve compatibility and decrease aggressivity in
the heaviest vehicles. Daimler emphasized that all vehicles will have
to comply with the Federal Motor Vehicle Safety Standards, and will
likely be designed to test well in NHTSA's NCAP tests.
Other commenters (Aluminum Association, CARB, CAS, ICCT, MEMA,
NRDC, U.S. Steel) also emphasized the need for NHTSA to account for the
safety benefits to be expected in the future from use of advanced
materials for lightweighting purposes and other engineering advances.
The Aluminum Association stated that advanced vehicle design and
construction techniques using aluminum can improve energy management
and minimize adverse safety effects of their use,\123\ but that NHTSA's
safety analysis could not account for those benefits if it were based
on MYs 1991-1999 vehicles. CAS, ICCT, and U.S. Steel discussed similar
benefits for more recent and future vehicles built with high strength
steel (HSS), although U.S. Steel cautioned that given the stringency of
the proposed standards, manufacturers would likely be encouraged to
build smaller and lighter vehicles in order to achieve compliance,
which fare worse in head-on collisions than larger, heavier vehicles.
AISI, in contrast to U.S. Steel, stated that in its research with the
Auto/Steel Partnership and in programs supported by DOE, it had found
that the use of new Advanced HSS steel grades could enable mass of
critical crash structures, such as front rails and bumper systems, to
be reduced by 25 percent without degrading performance in standard
NHTSA frontal or IIHS offset

[[Page 25387]]

instrumented crash tests compared to their ``heavier counterparts.''
---------------------------------------------------------------------------

\123\ The Aluminum Association (NHTSA-2009-0059-0067.3) stated
that its research on vehicle safety compatibility between an SUV and
a mid-sized car, done jointly with DRI, shows that reducing the
weight of a heavier SUV by 20% (a realistic value for an aluminum-
intensive vehicle) could reduce the combined injury rate for both
vehicles by 28% in moderately severe crashes. The commenter stated
that it would keep NHTSA apprised of its results as its research
progressed. Based on the information presented, NHTSA believes that
this research appears to agree with NHTSA's latest analysis, which
finds that a reduction in weight for the heaviest vehicles may
improve overall fleet safety.
---------------------------------------------------------------------------

Agencies' response: NHTSA, in consultation with EPA and DOE, plans
to begin updating the MYs 1991-1999 database on which NHTSA's safety
analyses in the NPRM and final rule are based in the next several
months in order to analyze the differences in safety effects against
vehicles built in more recent model years. As this task will take at
least a year to complete, beginning it immediately after the NPRM would
not have enabled the agency to complete it and then conduct a new
analysis during the period between the NPRM and the final rule.
For purposes of this final rule, however, we believe that using the
same MYs 1991-1999 database as that used in the 2003 Kahane study
provides a reasonable basis for attempting to estimate safety effects
due to reductions in mass. While commenters often stated that updating
the database would help to reveal the effect of recently-introduced
lightweight vehicles with extensive material substitution, there have
in fact not yet been a significant number of vehicles with substantial
mass reduction/material substitution to analyze, and they must also
show up in the crash databases for NHTSA to be able to add them to its
analysis. Based on NHTSA's research, specifically, on three statistical
analyses over a 12-year period (1991-2003) covering a range of 22 model
years (1978-1999), NHTSA believes that the relationships between mass,
size, and safety has only changed slowly over time, although we
recognize that they may change somewhat more rapidly in the
future.\124\ As the on-road fleet gains increasing numbers of vehicles
with increasing amounts of different methods of mass reduction applied
to them, we may begin to discern changes in the crash databases due to
the presence of these vehicles, but any such changes are likely to be
slow and evolutionary, particularly in the context of MYs 2000-2009
vehicles. The agencies do expect that further analysis of historical
data files will continue to provide a robust and practicable basis for
estimating the potential safety effects that might occur with future
reductions in vehicle mass. However, we recognize that estimates
derived from analysis of historical data, like estimates from any other
type of analysis (including simulation-based analysis, which cannot
feasibly cover all relevant scenarios), will be uncertain in terms of
predicting actual future outcomes with respect to a vehicle fleet,
driving population, and operating environment that does not yet exist.
---------------------------------------------------------------------------

\124\ NHTSA notes the CAS' comments regarding changes in the
vehicle fleets since the introduction of CAFE standards in the late
1970s, but believes they apply more to the differences between late
1970s through 1980s vehicles and 2010s vehicles than to the
differences between 1990s and 2010s vehicles. NHTSA believes that
the CAS comments regarding the phase-out of 1970s vehicles and their
replacement with safer, better fuel-economy-achieving 1980s vehicles
paint with rather too large a brush to be relevant to the main
discussion of whether the 2003 Kahane report database can reasonably
be used to estimate safety effects of mass reduction for the MYs
2012-2016 fleet.
---------------------------------------------------------------------------

The agencies also recognize that more recent vehicles have more
safety features than 1990s vehicles, which are likely to make them
safer overall. To account for this, NHTSA did adjust the results of
both its NPRM and final rule analysis to include known safety
improvements, like ESC and increases in seat belt use, that have
occurred since MYs 1991-1999.\125\ However, simply because newer
vehicles have more safety countermeasures, does not mean that the
weight/safety relationship necessarily changes. More likely, it would
change the target population (the number of fatalities) to which one
would apply the weight/safety relationship. Thus, we still believe that
some mass reduction techniques for both passenger cars and light trucks
can make them less safe, in certain crashes as discussed in NHTSA's
FRIA, than if mass had not been reduced.\126\
---------------------------------------------------------------------------

\125\ See NHTSA FRIA Chapter IX.
\126\ If one has a vehicle (vehicle A), and both reduces the
vehicle's mass and adds new safety equipment to it, thus creating a
variant (vehicle A1), the variant might conceivably have
a level of overall safety for its occupants equal to that of the
original vehicle (vehicle A). However, vehicle A1 might
not be as safe as second variant (vehicle A2) of vehicle
A, one that is produced by adding to vehicle A the same new safety
equipment added to the first variant, but this time without any mass
reduction.
---------------------------------------------------------------------------

As for NHTSA's assumptions about mass reduction, in its analysis,
NHTSA generally assumed that lighter vehicles could be reduced in
weight by 5 percent while heavier light trucks could be reduced in
weight by 10 percent. NHTSA recognizes that manufacturers might choose
a different mass reduction scheme than this, and that its
quantification of the estimated effect on safety would be different if
they did. We emphasize that our estimates are based on the assumptions
we have employed and are intended to help the agency consider the
potential effect of the final standards on vehicle safety. Thus, based
on the 2010 Kahane analysis, reductions in weight for the heavier light
trucks would have positive overall safety effects,\127\ while mass
reductions for passenger cars and smaller light trucks would have
negative overall safety effects.
---------------------------------------------------------------------------

\127\ This is due to the beneficial effect on the occupants of
vehicles struck by the downweighted larger vehicles.

NHTSA's estimates are inaccurate because they do not account for the
fact that lighter vehicles are safer than heavier cars in terms of
---------------------------------------------------------------------------
crash-avoidance, because they handle and brake better

ICCT stated that lighter vehicles are better able to avoid crashes
because they ``handle and brake slightly better,'' arguing that size-
based standards encourage lighter-weight car-based SUVs with
``significantly better handling and crash protection'' than 1996-1999
mid-size SUVs, which will reduce both fatalities and fuel consumption.
ICCT stated that NHTSA did not include these safety benefits in its
analysis. DRI also stated that its 2005 report found that crash
avoidance improves with reduction in curb weight and/or with increases
in wheelbase and track, because ``Crash avoidance can depend, amongst
other factors, on the vehicle directional control and rollover
characteristics.'' DRI argued that, therefore, ``These results indicate
that vehicle weight reduction tends to decrease fatalities, but vehicle
wheelbase and track reduction tends to increase fatalities.''
Agencies' response: In fact, NHTSA's regression analysis of crash
fatalities per million registration years measures the effects of crash
avoidance, if there are any, as well as crashworthiness. Given that the
historical empirical data for passenger cars show a trend of higher
crash rates for lighter cars, it is unclear whether lighter cars have,
in the net, superior crash avoidance, although the agencies recognize
that they may have advantages in certain individual situations. EPA
presents a discussion of improved accident avoidance as vehicle mass is
reduced in Chapter 7.6 of its final RIA. The important point to
emphasize is that it depends on the situation--it would oversimplify
drastically to point to one situation in which extra mass helps or
hurts and then extrapolate effects for crash avoidance across the board
based on only that.
For example, the relationship of vehicle mass to rollover and
directional stability is more complex than commenters imply. For
rollover, it is true that if heavy pickups were always more top-heavy
than lighter pickups of the same footprint, their higher center of
gravity could make them more rollover-prone, yet some mass can be
placed so as to lower a vehicle's center of gravity and make it less
rollover-prone. For mass reduction to be beneficial in rollover
crashes, then, it must take

[[Page 25388]]

center of gravity height into account along with other factors such as
passenger compartment design and structure, suspension, the presence of
various safety equipment, and so forth.
Similarly, for directional stability, it is true that having more
mass increases the ``understeer gradient'' of cars--i.e., it reinforces
their tendency to proceed in a straight line and slows their response
to steering input, which would be harmful where prompt steering
response is essential, such as in a double-lane-change maneuver to
avoid an obstacle. Yet more mass and a higher understeer gradient could
help when it is better to remain on a straight path, such as on a
straight road with icy patches where wheel slip might impair
directional stability. Thus, while less vehicle mass can sometimes
improve crash avoidance capability, there can also be situations when
more vehicle mass can help in other kinds of crash avoidance.
Further, NHTSA's research suggests that additional vehicle mass may
be even more helpful, as discussed in Chapter IX of NHTSA's FRIA, when
the average driver's response to a vehicle's maneuverability is taken
into account. Lighter cars have historically (1976-2009) had higher
collision-involvement rates than heavier cars--even in multi-vehicle
crashes where directional and rollover stability is not particularly an
issue.\128\ Based on our analyses using nationally-collected FARS and
GES data, drivers of lighter cars are more likely to be the culpable
party in a 2-vehicle collision, even after controlling for footprint,
the driver's age, gender, urbanization, and region of the country.
---------------------------------------------------------------------------

\128\ See, e.g., NHTSA (2000). Traffic Safety Facts 1999. Report
No. DOT HS 809 100. Washington, DC: National Highway Traffic Safety
Administration, p. 71; Najm, W.G., Sen, B., Smith, J.D., and
Campbell, B.N. (2003). Analysis of Light Vehicle Crashes and Pre-
Crash Scenarios Based on the 2000 General Estimates System, Report
No. DOT HS 809 573. Washington, DC: National Highway Traffic Safety
Administration, p. 48.
---------------------------------------------------------------------------

Thus, based on this data, it appears that lighter cars may not be
driven as well as heavier cars, although it is unknown why this is so.
If poor drivers intrinsically chose light cars (self-selection), it
might be evidenced by an increase in antisocial driving behavior (such
as DWI, drug involvement, speeding, or driving without a license) as
car weight decreases, after controlling for driver age and gender--in
addition to the increases in merely culpable driver behavior (such as
failure to yield the right of way). But analyses in NHTSA's 2003 report
did not show an increase in antisocial driver behavior in the lighter
cars paralleling their increase in culpable involvements.
NHTSA also hypothesizes that certain aspects of lightness and/or
smallness in a car may give a driver a perception of greater
maneuverability that ultimately results in driving with less of a
``safety margin,'' e.g., encouraging them to weave in traffic. That may
appear paradoxical at first glance, as maneuverability is, in the
abstract, a safety plus. Yet the situation is not unlike powerful
engines that could theoretically enable a driver to escape some
hazards, but in reality have long been associated with high crash and
fatality rates.\129\
---------------------------------------------------------------------------

\129\ Robertson, L.S. (1991), ``How to Save Fuel and Reduce
Injuries in Automobiles,'' The Journal of Trauma, Vol. 31, pp. 107-
109; Kahane, C.J. (1994). Correlation of NCAP Performance with
Fatality Risk in Actual Head-On Collisions, NHTSA Technical Report
No. DOT HS 808 061. Washington, DC: National Highway Traffic Safety
Administration, http://www-nrd.nhtsa.dot.gov/Pubs/808061.PDF, pp. 4-
7.

NHTSA's estimates are inaccurate because fatalities are linked more to
---------------------------------------------------------------------------
other factors than mass

Tom Wenzel stated that the safety record of recent model year
crossover SUVs indicates that weight reduction in this class of
vehicles (small to mid-size SUVs) resulted in a reduction in fatality
risk. Wenzel argued that NHTSA should acknowledge that other vehicle
attributes may be as important, if not more important, than vehicle
weight or footprint in terms of occupant safety, such as unibody
construction as compared to ladder-frame, lower bumpers, and less rigid
frontal structures, all of which make crossover SUVs more compatible
with cars than truck-based SUVs.
Marc Ross commented that fatalities are linked more strongly to
intrusion than to mass, and stated that research by safety experts in
Japan and Europe suggests the main cause of serious injuries and deaths
is intrusion due to the failure of load-bearing elements to properly
protect occupants in a severe crash. Ross argued that the results from
this project have ``overturned the original views about
compatibility,'' which thought that mass and the mass ratio were the
dominant factors. Since footprint-based standards will encourage the
reduction of vehicle weight through materials substitution while
maintaining size, Ross stated, they will help to reduce intrusion and
consequently fatalities, as the lower weight reduces crash forces while
maintaining size preserves crush space. Ross argued that this factor
was not considered by NHTSA in its discussion of safety. ICCT agreed
with Ross' comments on this issue.
In previous comments on NHTSA rulemakings and in several studies,
Wenzel and Ross have argued generally that vehicle design and
``quality'' is a much more important determinant of vehicle safety than
mass. In comments on the NPRM, CARB, NRDC, Sierra Club, and UCS echoed
this theme.
ICCT commented as well that fatality rates in the EU are much lower
than rates in the U.S., even though the vehicles in the EU fleet tend
to be smaller and lighter than those in the U.S. fleet. Thus, ICCT
argued, ``This strongly supports the idea that vehicle and highway
design are far more important factors than size or weight in vehicle
safety.'' ICCT added that ``It also suggests that the rise in SUVs in
the U.S. has not helped reduce fatalities.'' CAS also commented that
Germany's vehicle fleet is both smaller and lighter than the American
fleet, and has lower fatality rates.
Agencies' response: NHTSA and EPA agree that there are many
features that affect safety. While crossover SUVs have lower fatality
rates than truck-based SUVs, there are no analyses that attribute the
improved safety to mass alone, and not to other factors such as the
lower center of gravity or the unibody construction of these vehicles.
While a number of improvements in safety can be made, they do not
negate the potential that another 100 lbs. could make a passenger car
or crossover vehicle safer for its occupants, because of the effects of
mass per se as discussed in NHTSA's FRIA, albeit similar mass
reductions could make heavier LTVs safer to other vehicles without
necessarily harming their own drivers and occupants. Moreover, in the
2004 response to docket comments, NHTSA explained that the significant
relationship between mass and fatality risk persisted even after
controlling for vehicle price or nameplate, suggesting that vehicle
``quality'' as cited by Wenzel and Ross is not necessarily more
important than vehicle mass.
As for reductions in intrusions due to material substitution, the
agencies agree generally that the use of new and innovative materials
may have the potential to reduce crash fatalities, but such vehicles
have not been introduced in large numbers into the vehicle fleet. The
agencies will continue to monitor the situation, but ultimately the
effects of different methods of mass reduction on overall safety in the
real world (not just in simulations) will need to be analyzed when
vehicles with these types of mass reduction are on the road in
sufficient quantities to provide statistically significant results. For
example, a vehicle that is designed to be

[[Page 25389]]

much stiffer to reduce intrusion is likely to have a more severe crash
pulse and thus impose greater forces on the occupants during a crash,
and might not necessarily be good for elderly and child occupant safety
in certain types of crashes. Such trade-offs make it difficult to
estimate overall results accurately without real world data. The
agencies will continue to evaluate and analyze such real world data as
it becomes available, and will keep the public informed as to our
progress.
ICCT's comment illustrates the fact that different vehicle fleets
in different countries can face different challenges. NHTSA does not
believe that the fact that the EU vehicle fleet is generally lighter
than the U.S. fleet is the exclusive reason, or even the primary
factor, for the EU's lower fatality rates. The data ICCT cites do not
account for significant differences between the U.S. and EU such as in
belt usage, drunk driving, rural/urban roads, driving culture, etc.

The structure of the standards reduces/contributes to potential safety
risks from mass reduction

Since switching in 2006 to setting attribute-based light truck CAFE
standards, NHTSA has emphasized that one of the benefits of a
footprint-based standard is that it discourages manufacturers from
building smaller, less safe vehicles to achieve CAFE compliance by
``balancing out'' their larger vehicles, and thus avoids a negative
safety consequence of increasing CAFE stringency.\130\ Some commenters
on the NPRM (Daimler, IIHS, NADA, NRDC, Sierra Club et al.) agreed that
footprint-based standards would protect against downsizing and help to
mitigate safety risks, while others stated that there would still be
safety risks even with footprint-based standards--CEI, for example,
argued that mass reduction inherently creates safety risks, while IIHS
and Porsche expressed concern about footprint-based standards
encouraging manufacturers to manipulate wheelbase, which could reduce
crush space and worsen vehicle handling. U.S. Steel and AISI both
commented that the ``aggressive schedule'' for the proposed increases
in stringency could encourage manufacturers to build smaller, lighter
vehicles in order to comply.
---------------------------------------------------------------------------

\130\ We note that commenters were divided on whether they
believed there was a clear correlation between vehicle size/weight
and safety (CEI, Congress of Racial Equality, Heritage Foundation,
IIHS, Spurgeon, University of PA Environmental Law Project) or
whether they believed that the correlation was less clear, for
example, because they believed that vehicle design was more
important than vehicle mass (CARB, Public Citizen).
---------------------------------------------------------------------------

Some commenters also focused on the shape and stringency of the
target curves and their potential effect on vehicle safety. IIHS agreed
with the agencies' tentative decision to cut off the target curves at
the small-footprint end. Regarding the safety effect of the curves
requiring less stringent targets for larger vehicles, while IIHS stated
that increasing footprint is good for safety, CAS, Wenzel, and the UCSB
students stated that decreasing footprint may be better for safety in
terms of risk to occupants of other vehicles. Daimler, Wenzel, and the
University of PA Environmental Law Project commented generally that
more similar passenger car and light truck targets at identical
footprints (as Wenzel put it, a single target curve) would improve
fleet compatibility and thus, safety, by encouraging manufacturers to
build more passenger cars instead of light trucks.
Agencies' response: The agencies continue to believe that
footprint-based standards help to mitigate potential safety risks from
downsizing if the target curves maintain sufficient slope, because,
based on NHTSA's analysis, larger-footprint vehicles are safer than
smaller-footprint vehicles.\131\ The structure of the footprint-based
curves will also discourage the upsizing of vehicles. Nevertheless, we
recognize that footprint-based standards are not a panacea--NHTSA's
analysis continues to show that there was a historical relationship
between lower vehicle mass and increased safety risk in passenger cars
even if footprint is maintained, and there are ways that manufacturers
may increase footprint that either improve or reduce vehicle safety, as
indicated by IIHS and Porsche.
---------------------------------------------------------------------------

\131\ See Chapter IX of NHTSA's FRIA.
---------------------------------------------------------------------------

With regard to whether the agencies should set separate curves or a
single one, NHTSA also notes in Section II.C that EPCA requires NHTSA
to establish standards separately for passenger cars and light trucks,
and thus concludes that the standards for each fleet should be based on
the characteristics of vehicles in each fleet. In other words, the
passenger car curve should be based on the characteristics of passenger
cars, and the light truck curve should be based on the characteristics
of light trucks--thus to the extent that those characteristics are
different, an artificially-forced convergence would not accurately
reflect those differences. However, such convergence could be
appropriate depending on future trends in the light vehicle market,
specifically further reduction in the differences between passenger car
and light truck characteristics. While that trend was more apparent
when car-like 2WD SUVs were classified as light trucks, it seems likely
to diminish for the model year vehicles subject to these rules as the
truck fleet will be more purely ``truck-like'' than has been the case
in recent years.

NHTSA's estimates are inaccurate because NHTSA could mitigate
additional safety risks from mass reduction, if there are any, by
simply regulating safety more

Since NHTSA began considering the potential safety risks from mass
reduction in response to increased CAFE standards, some commenters have
suggested that NHTSA could mitigate those safety risks, if any, by
simply regulating more.\132\ In response to the safety analysis
presented in the NPRM, several commenters stated that NHTSA should
develop additional safety regulations to require vehicles to be
designed more safely, whether to improve compatibility (Adcock, NY DEC,
Public Citizen, UCS), to require seat belt use (CAS, UCS), to improve
rollover and roof crush resistance (UCS), or to improve crashworthiness
generally by strengthening NCAP and the star rating system (Adcock).
Wenzel commented further that ``Improvements in safety regulations will
have a greater effect on occupant safety than FE standards that are
structured to maintain, but may actually increase, vehicle size.''
---------------------------------------------------------------------------

\132\ See, e.g., MY 2011 CAFE final rule, 74 FR 14403-05 (Mar.
30, 2009).
---------------------------------------------------------------------------

Agencies' response: NHTSA appreciates the commenters' suggestions
and notes that the agency is continually striving to improve motor
vehicle safety consistent with its mission. As noted above, improving
safety in other areas affects the target population that the mass/
footprint relationship could affect, but it does not necessarily change
the relationship.
The 2010 Kahane analysis discussed in this final rule evaluates the
relative safety risk when vehicles are made lighter than they might
otherwise be absent the final MYs 2012-2016 standards. It does consider
the effect of known safety regulations as they are projected to affect
the target population.

Casualty risks range widely for vehicles of the same weight or
footprint, which skews regression analysis and makes computer
simulation a better predictor of the safety effects of mass reduction

[[Page 25390]]

Wenzel commented that he had found, in his most recent work, after
accounting for drivers and crash location, that there is a wide range
in casualty risk for vehicles with the same weight or footprint. Wenzel
stated that for drivers, casualty risk does generally decrease as
weight or footprint increases, especially for passenger cars, but the
degree of variation in the data for vehicles (particularly light
trucks) at a given weight or footprint makes it difficult to say that a
decrease in weight or footprint will necessarily result in increased
casualty risk. In terms of risk imposed on the drivers of other
vehicles, Wenzel stated that risk increases as light truck weight or
footprint increases.
Wenzel further stated that because a regression analysis can only
consider the average trend in the relationship between vehicle weight/
size and risk, it must ``ignore'' vehicles that do not follow that
trend. Wenzel therefore recommended that the agency employ computer
crash simulations for analyzing the effect of vehicle weight reduction
on safety, because they can ``pinpoint the effect of specific vehicle
designs on safety,'' and can model future vehicles which do not yet
exist and are not bound to analyzing historical data. Wenzel cited, as
an example, a DRI simulation study commissioned by the Aluminum
Association (Kebschull 2004), which used a computer model to simulate
the effect of changing SUV mass or footprint (without changing other
attributes of the vehicle) on crash outcomes, and showed a 15 percent
net decrease in injuries, while increasing wheelbase by 4.5 inches
while maintaining weight showed a 26 percent net decrease in serious
injuries.
Agencies' response: The agencies have reviewed Mr. Wenzel's draft
report for DOE to which he referred in his comments, but based on
NHTSA's work do not find such a wide range of safety risk for vehicles
with the same weight, although we agree there is a range of risk for a
given footprint. Wenzel found that for drivers, casualty risk does
generally decrease as weight or footprint increases, especially for
passenger cars, and that in terms of risk imposed on the drivers of
other vehicles, risk increases as light truck weight or footprint
increases, but concluded that the variation in the data precluded the
possibility of drawing any conclusions. In the 2010 Kahane study
presented in the FRIA, NHTSA undertook a similar analysis in which it
correlated weight to fatality risk for vehicles of essentially the same
footprint.\133\ The ``decile analysis,'' provided as a check on the
trend/direction of NHTSA's regression analysis, shows that societal
fatality risk generally increases and rarely decreases for lighter
relative to heavier cars of the same footprint. Thus, while Mr. Wenzel
was reluctant to draw a conclusion, NHTSA believes that both our
research and Mr. Wenzel's appear to point to the same conclusion. We
agree that there is a wide range in casualty risk among cars of the
same footprint, but we find that that casualty risk is correlated with
weight. The correlation shows that heavier cars have lower overall
societal fatality rates than lighter cars of very similar footprint.
---------------------------------------------------------------------------

\133\ Subsections 2.4 and 3.3 of new report.
---------------------------------------------------------------------------

The agencies agree that simulation can be beneficial in certain
circumstances. NHTSA cautions, however, that it is difficult for a
simulation analysis to capture the full range of variations in crash
situations in the way that a statistical regression analysis does.
Vehicle crash dynamics are complex, and small changes in initial crash
conditions (such as impact angle or closing speed) can have large
effects on injury outcome. This condition is a consequence of
variations in the deformation mode of individual components (e.g.,
buckling, bending, crushing, material failure, etc.) and how those
variations affect the creation and destruction of load paths between
the impacting object and the occupant compartment during the crash
event. It is therefore difficult to predict and assess structural
interactions using computational methods when one does not have a
detailed, as-built geometric and material model. Even when a complete
model is available, prudent engineering assessments require extensive
physical testing to verify crash behavior and safety. Despite all this,
the agencies recognize that detailed crash simulations can be useful in
estimating the relative structural effects of design changes over a
limited range of crash conditions, and will continue to evaluate the
appropriate use of this tool in the future.
Simplified crash simulations can also be valuable tools, but only
when employed as part of a comprehensive analytical program. They are
especially valuable in evaluating the relative effect and associated
confidence intervals of feasible design alternatives. For example, the
method employed by Nusholtz et al.\134\ could be used by a vehicle
designer to estimate the benefit of incremental changes in mass or
wheelbase as well as the tradeoffs that might be made between them once
that designer has settled on a preliminary design. A key difference
between the research by Nusholtz and the research by Kebschull that Mr.
Wenzel cited \135\ is in their suggested applications. The former is
useful in evaluating proposed alternatives early in the design
process--Nusholtz specifically warns that the model provides only
``general insights into the overall risk * * * and cannot be used to
obtain specific response characteristics.'' Mr. Wenzel implies the
latter can ``isolate the effect of specific design changes, such as
weight reduction'' and thus quantify the fleet-wide effect of
substantial vehicle redesigns. Yet while Kebschull reports injury
reductions to three significant digits, there is no validation that
vehicle structures of the proposed weight and stiffness are even
feasible with current technology. Thus, while the agencies agree that
computer simulations can be useful tools, we also recognize the value
of statistical regression analysis for determining fleet-wide effects,
because it inherently incorporates real-world factors in historical
safety assessments.
---------------------------------------------------------------------------

\134\ Nusholtz, G.S., G. Rabbiolo, and Y. Shi, ``Estimation of
the Effects of Vehicle Size and Mass on Crash-Injury Outcome Through
Parameterized Probability Manifolds,'' Society of Automotive
Engineers (2003), Document No. 2003-01-0905. Available at http://
www.sae.org/technical/papers/2003-01-0905 (last accessed Feb. 15,
2010).
\135\ Mr. Wenzel cites the report by Kebschull et al. [2004,
DRI-TR-04-04-02] as an example of what he regards as the effective
use of computer crash simulation. NHTSA does not concur that this
analysis represents a viable analytical method for evaluating the
fleet-wide tradeoffs between vehicle mass and societal safety. The
simulation method employed was not a full finite element
representation of each major structural component in the vehicles in
question. Instead, an Articulated Total Body (ATB) representation
was constructed for each of two representative vehicles. In the ATB
model, large structural subsystems were represented by a single
ellipsoid. Consolidated load-deflection properties of these
subsystems and the joints that tie them together were ``calibrated''
for an ATB vehicle model by requiring that it reproduce the
acceleration pulse of a physical NHTSA crash test. NHTSA notes that
vehicle simulation models that are calibrated to a single crash test
configuration (e.g., a longitudinal NCAP test into a rigid wall) are
often ill-equipped to analyze alternative crash scenarios (e.g.,
vehicle-to-vehicle crashes at arbitrary angles and lateral offsets).

DRI's analysis shows that lighter vehicles will save lives, and NHTSA
---------------------------------------------------------------------------
reaches the opposite conclusion without disproving DRI's analysis

The difference between NHTSA's results and DRI's results for the
relationship between vehicle mass and vehicle safety has been at the
crux of this issue for several years. While NHTSA offered some theories
in the NPRM as to why DRI might have found a safety benefit for mass
reduction, NHTSA's work since then has enabled it to identify what we
believe is the most likely reason for DRI's findings.

[[Page 25391]]

The potential near multicollinearity of the variables of curb weight,
track width, and wheelbase creates some degree of concern that any
regression models with those variables could inaccurately calibrate
their effects. However, based on its own experience with statistical
analysis, NHTSA believes that the specific two-step regression model
used by DRI increases this concern, because it weakens relationships
between curb weight and dependent variables by splitting the effect of
curb weight across the two regression steps.
The comments below are in response to NHTSA's theories in the NPRM
about the source of the differences between NHTSA's and DRI's results.
The majority of them are answered more fully in the 2010 Kahane report
included in NHTSA's FRIA, but we respond to them in this document as
well for purposes of completeness.

NHTSA and DRI may have reached different conclusions because NHTSA's
study does not distinguish between reductions in size and reductions in
weight like DRI's

Several commenters (CARB, CBD, EDF, ICCT, NRDC, and UCS) stated
that DRI had been able to separate the effect of size and weight in its
analysis, and in so doing proved that there was a safety benefit to
reducing weight without reducing size. The commenters suggested that if
NHTSA properly distinguished between reductions in size and reductions
in weight, it would find the same result as DRI.
Agencies' response: In the 2010 Kahane analysis presented in the
FRIA, NHTSA did attempt to separate the effects of vehicle size and
weight by performing regression analyses with footprint (or
alternatively track width and wheelbase) and curb weight as separate
independent variables. For passenger cars, NHTSA found that the
regressions attribute the fatality increase due to downsizing about
equally to mass and footprint--that is, the effect of reducing mass
alone is about half the effect of reducing mass and reducing footprint.
Unlike DRI's results, NHTSA's regressions for passenger cars and for
lighter LTVs did not find a safety benefit to reducing weight without
reducing size; while NHTSA did find a safety benefit for reducing
weight in the heaviest LTVs, the magnitude of the benefit as compared
to DRI's was significantly smaller. NHTSA believes that these
differences in results may be an artifact of DRI's two-step regression
model, as explained above.

NHTSA and DRI may have reached different conclusions because NHTSA's
study does not include two-door cars like DRI's

One of NHTSA's primary theories in the NPRM as to why NHTSA and
DRI's results differed related to DRI's inclusion in its analysis of 2-
door cars. NHTSA had excluded those vehicles from its analysis on the
grounds that 2-door cars had a disproportionate crash rate (perhaps due
to their inclusion of muscle and sports cars) which appeared likely to
skew the regression. Several commenters argued that NHTSA should have
included 2-door cars in its analysis. DRI and James Adcock stated that
2-door cars should not be excluded because they represent a significant
portion of the light-duty fleet, while CARB and ICCT stated that
because DRI found safety benefits whether 2-door cars were included or
not, NHTSA should include 2-door cars in its analysis. Wenzel also
commented that NHTSA should include 2-door cars in subsequent analyses,
stating that while his analysis of MY 2000-2004 crash data from 5
states indicates that, in general, 4-door cars tend to have lower
fatality risk than 2-door cars, the risk is even lower when he accounts
for driver age/gender and crash location. Wenzel suggested that the
increased fatality risk in the 2-door car population seemed primarily
attributable to the sports cars, and that that was not sufficient
grounds to exclude all 2-door cars from NHTSA's analysis.
Agencies' response: The agencies agree that 2-door cars can be
included in the analysis, and NHTSA retracts previous statements that
DRI's inclusion of them was incorrect. In its 2010 analysis, NHTSA
finds that it makes little difference to the results whether 2-door
cars are included, partially included, or excluded from the analysis.
Thus, analyses of 2-door and 4-door cars combined, as well as other
combinations, have been included in the analysis. That said, no
combination of 2-door and 4-door cars resulted in NHTSA's finding a
safety benefit for passenger cars due to mass reduction.

NHTSA and DRI may have reached different conclusions due to different
assumptions

DRI commented that the differences found between its study and
NHTSA's may be due to the different assumptions about the linearity of
the curb weight effect and control variable for driver age, vehicle
age, road conditions, and other factors. NHTSA's analysis was based on
a two-piece linear model for curb weight with two different weight
groups (less than 2,950 lbs., and greater than or equal to 2.950 lbs).
The DRI analysis assumed a linear model for curb weight with a single
weight group. Additionally, DRI stated that NHTSA's use of eight
control variables (rather than three control variables like DRI used)
for driver age introduces additional degrees of freedom into the
regressions, which it suggested may be correlated with the curb weight,
wheelbase, and track width, and/or other control variables. DRI
suggested that this may also affect the results and cause or contribute
to the differences in outcomes between NHTSA and DRI.
Agencies' response: NHTSA's FRIA documents that NHTSA analyzed its
database using both a single parameter for weight (a linear model) and
two parameters for weight (a two-piece linear model). In both cases,
the logistic regression responded identically, allocating the same way
between weight, wheelbase, track width, or footprint.\136\ Thus, NHTSA
does not believe that the differences between its results and DRI's
results are due to whether the studies used a single weight group or
two weight groups.
---------------------------------------------------------------------------

\136\ Subsections 2.2 and 2.3 of new report.
---------------------------------------------------------------------------

The FRIA also documents that NHTSA examined NHTSA's use of eight
control variables for driver age (ages 14-30, 30-50, 50-70, 70+ for
males and females separately, versus DRI's use of three control
variables for age (FEMALE = 1 for females, 0 for males, YOUNGDRV = 35-
AGE for drivers under 35, 0 for all others, OLDMAN = AGE-50 for males
over 50, 0 for all others; OLDWOMAN = AGE-45 for females over 45, 0 for
all others) to see if that affected the results. NHTSA ran its analysis
using the eight control variables and again using three control
variables for age, and obtained similar results each time.\137\ Thus,
NHTSA does not believe that the differences between its results and
DRI's results are due to the number of control variables used for
driver age.
---------------------------------------------------------------------------

\137\ Id.

NHTSA's and DRI's conclusions may be similar if confidence intervals
---------------------------------------------------------------------------
are taken into account

DRI commented that NHTSA has not reported confidence intervals,
while DRI has reported them in its studies. Thus, DRI argued, it is not
possible to determine whether the confidence intervals overlap and
whether the differences between NHTSA's and DRI's analyses are
statistically significant.
Agencies' response: NHTSA has included confidence intervals for the
main results of the 2010 Kahane analysis, as shown in Chapter IX of
NHTSA's FRIA. For passenger cars, the NHTSA results are a statistically

[[Page 25392]]

significant increase in fatalities with a 100 pound reduction while
maintaining track width and wheelbase (or footprint); the DRI results
are a statistically significant decrease in fatalities with a 100 pound
reduction while maintaining track width and wheelbase. The DRI results
are thus outside the confidence bounds of the NHTSA results and do not
overlap.

NHTSA should include a ``best-case'' estimate in its study

Several commenters (Center for Auto Safety, NRDC, Public Citizen,
Sierra Club et al., and Wenzel) urged NHTSA to include a ``best-case''
estimate in the final rule, showing scenarios in which lives were saved
rather than lost. Public Citizen stated that there would be safety
benefits to reducing the weight of the heaviest vehicles while leaving
the weight of the lighter vehicles unchanged, and that increasing the
number of smaller vehicles would provide safety benefits to
pedestrians, bicyclists, and motorcyclists. Sierra Club et al. stated
that new materials, smart design, and lighter, more advanced engines
can all improve fuel economy while maintaining or increasing vehicle
safety. Both Center for Auto Safety and Sierra Club argued that the
agency should have presented a ``best-case'' scenario to balance out
the ``worst-case'' scenario presented in the NPRM, especially if NHTSA
itself believed that the worst-case scenario was not inevitable. NRDC
requested that NHTSA present both a ``best-case'' and a ``most likely''
scenario. Wenzel simply stated that NHTSA did not present a ``best-
case'' scenario, despite DRI's finding in 2005 that fatalities would be
reduced if track width was held constant.
Agencies' response: NHTSA has included an ``upper estimate'' and a
``lower estimate'' in the new 2010 Kahane analysis. The lower estimate
assumes that mass reduction will be accomplished entirely by material
substitution or other techniques that do not perceptibly change a
vehicle's shape, structural strength, or ride quality. The lower
estimate examines specific crash modes and is meant to reflect the
increase in fatalities for the specific crash modes in which a
reduction in mass per se in the case vehicle would result in a
reduction in safety: namely, collisions with larger vehicles not
covered by the regulations (e.g., trucks with a GVWR over 10,000 lbs),
collisions with partially-movable objects (e.g., some trees, poles,
parked cars, etc.), and collisions of cars or light LTVs with heavier
LTVs--as well as the specific crash modes where a reduction in mass per
se in the case vehicle would benefit safety: namely, collisions of
heavy LTVs with cars or lighter LTVs. NHTSA believes that this is the
effect of mass per se, i.e., the effects of reduced mass will generally
persist in these crashes regardless of how the mass is reduced. The
lower estimate attempts to quantify that scenario, although any such
estimate is hypothetical and subject to considerable uncertainty. NHTSA
believes that a ``most likely'' scenario cannot be determined with any
certainty, and would depend entirely upon agency assumptions about how
manufacturers intend to reduce mass in their vehicles. While we can
speculate upon the potential effects of different methods of mass
reduction, we cannot predict with certainty what manufacturers will
ultimately do.

NHTSA should not include a ``worst-case'' estimate in its study

NRDC, Public Citizen and Sierra Club et al. commented that NHTSA
should remove the ``worst-case scenario'' estimate from the rulemaking,
generally because it was based on an analysis that evaluated historical
vehicles, and future vehicles would be sufficiently different to render
the ``worst-case scenario'' inapplicable.
Agencies' response: NHTSA stated in the NPRM that the ``worst-case
scenario'' addressed the effect of a kind of downsizing (i.e., mass
reduction accompanied by footprint reduction) that was not likely to be
a consequence of attribute-based CAFE standards, and that the agency
would refine its analysis of such a scenario for the final rule. NHTSA
has not used the ``worst-case scenario'' in the final rule. Instead, we
present three scenarios: the first is an estimate based directly on the
regression coefficients of weight reduction while maintaining footprint
in the statistical analyses of historical data. As discussed above,
presenting this scenario is possible because NHTSA attempted to
separate the effects of weight and footprint reduction in the new
analysis. However, even the new analysis of LTVs produced some
coefficients that NHTSA did not consider entirely plausible. NHTSA also
presents an ``upper estimate'' in which those coefficients for the LTVs
were adjusted based on additional analyses and expert opinion as a
safety agency and a ``lower estimate,'' which estimates the effect if
mass reduction is accomplished entirely by safety-conscious
technologies such as material substitution.
3. How has NHTSA refined its analysis for purposes of estimating the
potential safety effects of this Final Rule?
During the past months, NHTSA has extensively reviewed the
literature on vehicle mass, size, and fatality risk. NHTSA now agrees
with DRI and other commenters that it is essential to analyze the
effect of mass independently from the effects of size parameters such
as wheelbase, track width, or footprint--and that the NPRM's ``worst-
case'' scenario based on downsizing (in which weight, wheelbase, and
track width could all be changed) is not useful for that purpose. The
agency should instead provide estimates that better reflect the more
likely effect of the regulation--estimating the effect of mass
reduction that maintains footprint.
Yet it is more difficult to analyze multiple, independent
parameters than a single parameter (e.g., curb weight), because there
is a potential concern that the near multicollinearity of the
parameters--the strong, natural and historical correlation of mass and
size--can lead to inaccurate statistical estimates of their
effects.\138\ NHTSA has performed new statistical analyses of its
historical database of passenger cars, light trucks, and vans (LTVs)
from its 2003 report (now including also 2-door cars), assessing
relationships between fatality risk, mass, and footprint. They are
described in Subsections 2.2 (cars) and 3.2 (LTVs) of the 2010 Kahane
report presented in Chapter IX of the FRIA. While the potential
concerns associated with near multicollinearity are inherent in
regression analyses with multiple size/mass parameters, NHTSA believes
that the analysis approach in the 2010 Kahane report, namely a single-
step regression analysis, generally reduces those concerns \139\ and
models the trends in the historical data. The results differ
substantially from DRI's, based on a two-step regression analysis.
Subsections 2.3 and 2.4 of the 2010

[[Page 25393]]

Kahane report attempt to account for the differences primarily by
applying selected techniques from DRI's analyses to NHTSA's database.
---------------------------------------------------------------------------

\138\ Greene, W. H. (1993). Econometric Analysis, Second
Edition. New York: Macmillan Publishing Company, pp. 266-268;
Allison, P.D. (1999), Logistic Regression Using the SAS System.
Cary, NC: SAS Institute Inc., pp. 48-51. The report shows variance
inflation factor (VIF) scores in the 5-7 range for curb weight,
wheelbase, and track width (or, alternatively, curb weight and
footprint) in NHTSA's database, exceeding the 2.5 level where near
multicollinearity begins to become a concern in logistic regression
analyses.
\139\ NHTSA believes that, given the near multicollinearity of
the independent variables, the two-step regression augments the
possibility of estimating inaccurate coefficients for curb weight,
because it weakens relationships between curb weight and dependent
variables by splitting the effect of curb weight across the two
regression steps as discussed further in Subsection 2.3 of NHTSA's
report.
---------------------------------------------------------------------------

The statistical analyses--logistic regressions--of trends in MYs
1991-1999 vehicles generate one set of estimates of the possible
effects of reducing mass by 100 pounds while maintaining footprint.
While these effects might conceivably carry over to future mass
reductions, there are two reasons that future safety effects of mass
reduction could differ from projections from historical data:
The statistical analyses are ``cross-sectional'' analyses
that estimate the increase in fatality rates for vehicles weighing n-
100 pounds relative to vehicles weighing n pounds, across the spectrum
of vehicles on the road, from the lightest to the heaviest. They do not
directly compare the fatality rates for a specific make and model
before and after a 100-pound reduction from that model. Instead, they
use the differences across makes and models as a surrogate for the
effects of actual reductions within a specific model; those cross-
sectional differences could include trends that are statistically, but
not causally related to mass.
The manner in which mass changed across MY 1991-1999
vehicles might not be consistent with future mass reductions, due to
the availability of newer materials and design methods.

Therefore, Subsections 2.5 and 3.4 of the 2010 Kahane report supplement
those estimates with one or more scenarios in which some of the
logistic regression coefficients are replaced by numbers based on
additional analyses and NHTSA's judgment of the likely effect of mass
per se (the ability to transfer momentum to other vehicles or objects
in a collision) and of what trends in the historical data could be
avoided by current mass-reduction technologies such as materials
substitution. The various scenarios may be viewed as a plausible range
of point estimates for the effects of mass reduction while maintaining
footprint, but they should not be construed as upper and lower bounds.
Furthermore, being point estimates, they are themselves subject to
uncertainties, such as, for example, the sampling errors associated
with statistical analyses.
The principal findings and conclusions of the 2010 Kahane report
are as follows:
Passenger cars: This database with the one-step regression method
of the 2003 Kahane report estimates an increase of 700-800 fatalities
when curb weight is reduced by 100 pounds and footprint is reduced by
0.65 square feet (the historic average footprint reduction per 100-
pound mass reduction in cars). The regression attributes the fatality
increase about equally to curb weight and to footprint. The results are
approximately the same whether 2-door cars are fully included or
partially included in the analysis or whether only 4-door cars are
included (as in the 2003 report). Regressions by curb weight, track
width and wheelbase produce findings quite similar to the regressions
by curb weight and footprint, but the results with the single ``size''
variable, footprint, rather than the two variables, track width and
wheelbase vary even less with the inclusion or exclusion of 2-door
cars.
In Subsection 2.3 of the new report, a two-step regression method
that resembles (without exactly replicating) the approach by DRI, when
applied to the same (NHTSA's) crash and registration data, estimates a
large benefit when mass is reduced, offset by even larger fatality
increases when track width and wheelbase (or footprint) are reduced.
NHTSA believes that the benefit estimated by this method is inaccurate,
due to the potential concerns with the near multicollinearity of the
parameters (curb weight, track width, and wheelbase) \140\ even though
the analysis is theoretically unbiased.\141\ Almost any analysis
incorporating those parameters has a possibility of inaccurate
coefficients due to near multicollinearity; however, based on our own
experience with other regression analyses of crash data, NHTSA believes
a DRI-type two-step method augments the possibility of estimating
inaccurate coefficients for curb weight, because it weakens
relationships between curb weight and dependent variables by splitting
the effect of curb weight across the two regression steps.
---------------------------------------------------------------------------

\140\ As evidenced by VIF scores in the 5-7 range, exceeding the
2.5 level where near multicollinearity begins to become a concern in
logistic regression analyses.
\141\ Subsection 2.3 of the 2010 Kahane report attempts to
explain why the two-step method, when applied to NHTSA's 2003
database, produces results a lot like DRI's, but it does not claim
that DRI obtained its results from its own database for exactly
those reasons. NHTSA did not analyze DRI's database. The two-step
method is ``theoretically unbiased'' in the sense that it seeks to
estimate the same parameters as the one-step analysis.
---------------------------------------------------------------------------

In Subsection 2.4 of the new report, as a check on the results from
the regression methods, NHTSA also performed what we refer to as
``decile'' analyses: Simpler, tabular data analysis that compares
fatality rates of cars of different mass but similar footprint. Decile
analysis is not a precise tool because it does not control for
confounding factors such as driver age/gender or the specific type of
car, but it may be helpful in identifying the general directional trend
in the data when footprint is held constant and curb weight varies. The
decile analyses show that fatality risk in MY 1991-1999 cars generally
increased and rarely decreased for lighter relative to heavier cars of
the same footprint. They suggest that the historical, cross-sectional
trend was generally in the lighter [harr] more fatalities direction and
not in the opposite direction, as might be suggested by the regression
coefficients from the method that resembles DRI's approach.
The regression coefficients from NHTSA's one-step method suggest
that mass and footprint each accounted for about half the fatality
increase associated with downsizing in a cross-sectional analysis of
1991-1999 cars. They estimate the historical difference in societal
fatality rates (i.e., including fatalities to occupants of all the
vehicles involved in the collisions, plus any pedestrians) of cars of
different curb weights but the same footprint. They may be considered
an ``upper-estimate scenario'' of the effect of future mass reduction--
if it were accomplished in a manner that resembled the historical
cross-sectional trend--i.e., without any particular regard for safety
(other than not to reduce footprint).
However, NHTSA believes that future vehicle design is likely to
take advantage of safety-conscious technologies such as materials
substitution that can reduce mass without perceptibly changing a car's
shape or ride and maintain its structural strength. This could avoid
much of the risk associated with lighter and smaller vehicles in the
historical analyses, especially the historical trend toward higher
crash-involvement rates for lighter and smaller vehicles.\142\ It could
thereby shrink the added risk close to just the effects of mass per se
(the ability to transfer momentum to other vehicles or objects in a
collision). Subsection 2.5 of the 2010 Kahane report attempts to
quantify a ``lower-estimate scenario'' for the potential effect of mass
reduction achieved by safety-conscious technologies; the estimated
effects are substantially smaller than in the upper-

[[Page 25394]]

estimate scenario based directly on the regression results.
---------------------------------------------------------------------------

\142\ This is discussed in greater depth in Subsections 2.1 and
2.5 of the 2010 Kahane report. The historic trend toward higher
crash-involvement rates for lighter and smaller vehicles is
documented in IIHS Advisory No. 5, July 1988, http://www.iihs.org/
research/advisories/iihs_advisory_5.pdf; IIHS News Release,
February 24, 1998, http://www.iihs.org/news/1998/iihs_news_
022498.pdf; Auto Insurance Loss Facts, September 2009, http://
www.iihs.org/research/hldi/fact_sheets/CollisionLoss_0909.pdf.
---------------------------------------------------------------------------

We note, again, that the preceding paragraph is conditional.
Nothing in the CAFE standard requires manufacturers to use material
substitution or, more generally, take a safety-conscious approach to
mass reduction.\143\ Federal Motor Vehicle Safety Standards include
performance tests that verify historical improvements in structural
strength and crashworthiness, but few FMVSS provide test information
that sheds light about how a vehicle rides or otherwise helps explain
the trend toward higher crash-involvement rates for lighter and smaller
vehicles. It is possible that using material substitution and other
current mass reduction methods could avoid the historical trend in this
area, but that remains to be studied as manufacturers introduce more of
these vehicles into the on-road fleet in coming years. A detailed
discussion of methods currently used for reducing the mass of passenger
cars and light trucks is included in Chapter 3 of the Technical Support
Document.
---------------------------------------------------------------------------

\143\ Footprint-based standards do not specify how or where to
remove mass while maintaining footprint, nor do they categorically
forbid footprint reductions, even if they discourage them.
---------------------------------------------------------------------------

LTVs: The principal difference between LTVs and passenger cars is
that mass reduction in the heavier LTVs is estimated to have
significant societal benefits, in that it reduces the fatality risk for
the occupants of cars and light LTVs that collide with the heavier
LTVs. By contrast, footprint (size) reduction in LTVs has a harmful
effect (for the LTVs' own occupants), as in cars. The regression method
of the 2003 Kahane report applied to the database of that report
estimates a societal increase of 231 fatalities when curb weight is
reduced by 100 pounds and footprint is reduced by 0.975 square feet
(the historic average footprint reduction per 100-pound mass reduction
in LTVs). But the regressions attribute an overall reduction of 266
fatalities to the 100-pound mass reduction and an increase of 497
fatalities to the .975-square-foot footprint reduction. The regression
results constitute one of the scenarios for the possible societal
effects of future mass reduction in LTVs.
However, NHTSA cautions that some of the regression coefficients,
even by NHTSA's preferred method, might not accurately model the
historical trend in the data, possibly due to near multicollinearity of
curb weight and footprint or because of the interaction of both of
these variables with LTV type.\144\ Based on supplementary analyses and
discussion in Subsections 3.3 and 3.4, the new report defines an
additional upper-estimate scenario that NHTSA believes may more
accurately reflect the historical trend in the data and a lower-
estimate scenario that may come closer to the effects of mass per se.
All three scenarios, however, attribute a societal fatality reduction
to mass reduction in the heavier LTVs.
---------------------------------------------------------------------------

\144\ For example, mid-size SUVs of the 1990s typically had high
mass relative to their short wheelbase and footprint (and
exceptionally high rates of fatal rollovers); minivans typically
have low mass relative to their footprint (and low fatality rates);
heavy-duty pickup trucks used extensively for work tend to have more
mass, for the same footprint, as basic full-sized pickup trucks that
are more often used for personal transportation.
---------------------------------------------------------------------------

Overall effects of mass reduction while maintaining footprint in
cars and LTVs: The immediate purpose of the new report's analyses of
relationships between fatality risk, mass, and footprint is to develop
the four parameters that the Volpe model needs in order to predict the
safety effects, if any, of the modeled mass reductions in MYs 2012-2016
cars and LTVs over the lifetime of those vehicles. The four numbers are
the overall percentage increases or decreases, per 100-pound mass
reduction while holding footprint constant, in crash fatalities
involving: (1) Cars < 2,950 pounds (which was the median curb weight of
cars in MY 1991-1999), (2) cars >= 2,950 pounds, (3) LTVs < 3,870
pounds (which was the median curb weight of LTVs in those model years),
and (4) LTVs >= 3,870 pounds. Here are the percentage effects for each
of the three alternative scenarios, again, the ``upper-estimate
scenario'' and the ``lower-estimate scenario'' have been developed
based on NHTSA's expert opinion as a vehicle safety agency:

Fatality Increase per 100-Pound Reduction (%) \145\
----------------------------------------------------------------------------------------------------------------
NHTSA expert
Actual regression opinion upper- NHTSA expert
result scenario estimate scenario opinion lower-
\146\ estimate scenario
----------------------------------------------------------------------------------------------------------------
Cars < 2,950 pounds.................................... 2.21 2.21 1.02
Cars >= 2,950 pounds................................... 0.90 0.90 0.44
LTVs < 3,870 pounds.................................... 0.17 0.55 0.41
LTVs >= 3,870 pounds................................... -1.90 -0.62 -0.73
----------------------------------------------------------------------------------------------------------------

In all three scenarios, the estimated effects of a 100-pound mass
reduction while maintaining footprint are an increase in fatalities in
cars < 2,950 pounds, substantially smaller increases in cars >= 2,950
pounds and LTVs < 3,870 pounds, and a societal benefit for LTVs >=
3,870 pounds (because it reduces fatality risk to occupants of cars and
lighter LTVs they collide with). These are the estimated effects of
reducing each vehicle by exactly 100 pounds. However, the actual mass
reduction will vary by make, model, and year. The aggregate effect on
fatalities can only be estimated by attempting to forecast, as NHTSA
has using inputs to the Volpe model, the mass reductions by make and
model. It should be noted, however, that a 100-pound reduction would be
5 percent of the mass of a 2000-pound car but only 2 percent of a 5000-
pound LTV. Thus, a forecast that mass will decrease by an equal or
greater percentage in the heavier vehicles than in the lightest cars
would be proportionately more influenced by the benefit for mass
reduction in the heavy LTVs than by the fatality increases in the other
groups; it is likely to result in an estimated net benefit under one or
more of the scenarios. It should also be noted, again, that the

[[Page 25395]]

three scenarios are point estimates and are subject to uncertainties,
such as the sampling errors associated with the regression results. In
the scenario based on actual regression results, the 1.96-sigma
sampling errors in the above estimates are 0.91 percentage
points for cars < 2,950 pounds and also for cars >= 2,950 pounds,
0.82 percentage points for LTVs < 3,870 pounds, and 1.18 percentage points for LTVs >= 3,870 pounds. In other words,
the fatality increase in the cars < 2,950 pounds and the societal
fatality reduction attributed to mass reduction in the LTVs >= 3,870
pounds are statistically significant. The sampling errors associated
with the scenario based on actual regression results perhaps also
indicate the general level of statistical noise in the other two
scenarios.
---------------------------------------------------------------------------

\145\ Reducing mass by 100 pounds in these vehicles is estimated
to have the listed percentage effect on fatalities in crashes
involving these vehicles. For example, if these vehicles are
involved in crashes that result in 10,000 fatalities, 2.21 means
that if mass is reduced by 100 pounds, fatalities will increase to
10,221 and -0.73 means fatalities will decrease to 9,927. In the
scenario based on actual regression results, the 1.96-sigma sampling
errors in the above estimates are 0.91 percentage points
for cars < 2,950 pounds and also for cars >= 2,950 pounds, 0.82 percentage points for LTVs < 3,870 pounds, and 1.18 percentage points for LTVs >= 3,870 pounds. In other
words, the fatality increase in the cars < 2,950 pounds and the
societal fatality reduction attributed to mass reduction in the LTVs
>= 3,870 pounds are statistically significant. The sampling errors
associated with the scenario based on actual regression results
perhaps also indicate the general level of statistical noise in the
other two scenarios.
\146\ For passenger cars, the upper-estimate scenario is the
actual-regression-result scenario.
---------------------------------------------------------------------------

4. What are the estimated safety effects of this Final Rule?
The table below shows the estimated safety effects of the modeled
reduction in vehicle mass provided in the NPRM and in this final rule
in order to meet the MYs 2012-2016 standards, based on the analysis
described briefly above and in much more detail in Chapter IX of the
FRIA. These are combined results for passenger cars and light trucks. A
positive number is an estimated increase in fatalities and a negative
number (shown in parentheses) is an estimated reduction in fatalities
over the lifetime of the model year vehicles compared to the MY 2011
baseline fleet.

----------------------------------------------------------------------------------------------------------------
MY 2012 MY 2013 MY 2014 MY 2015 MY 2016
----------------------------------------------------------------------------------------------------------------
NPRM ``Worst Case''.......... 34 54 194 313 493
NHTSA Expert Opinion Final 9 14 26 24 22
Rule Upper Estimate.........
NHTSA Expert Opinion Final 2 4 (17) (53) (80)
Rule Lower Estimate.........
Actual Regression Result 0 2 (94) (206) (301)
Scenario....................
----------------------------------------------------------------------------------------------------------------

NHTSA emphasizes that the table above is based on the NHTSA's
assumptions about how manufacturers might choose to reduce the mass of
their vehicles in response to the final rule, which are very similar to
EPA's assumptions. In general, as discussed above, the agencies assume
that mass will be reduced by as much as 10 percent in the heaviest LTVs
but only by as much as 5 percent in other vehicles and that substantial
mass reductions will take place only in the year that models are
redesigned. The actual mass reduction that is likely to occur in
response to the standards will of course vary by make and model,
depending on each manufacturer's particular approach, with likely more
opportunity for the largest LTVs that still use separate frame
construction.
The ``upper estimate'' presented above, as discussed in the FRIA,
assumes only that manufacturers will reduce vehicle mass without
reducing footprint. Thus, under such a scenario, safety effects could
be somewhat adverse if, for example, manufacturers chose to reduce
crush space associated with vehicle overhang as a way of reducing mass
without changing footprint. The ``lower estimate,'' in turn, is based
on the assumption that manufacturers will reduce vehicle mass solely
through methods like material substitution, which (under these
assumptions) fully maintain not only footprint but also all structural
integrity, and other aspects of vehicle safety. Under these scenarios,
safety effects could be worse if mass reduction was not undertaken
thoughtfully to maintain existing safety levels, but could also be
better if it was undertaken with a thorough and extensive vehicle
redesign to maximize both mass reduction and safety.
And finally, while NHTSA does not believe that the ``worst-case''
scenario presented in the NPRM is likely to occur during the MYs 2012-
2016 timeframe, we cannot guarantee that manufacturers will never
choose to reduce vehicle footprint, particularly if market forces lead
to increased sales of small vehicles in response to sharp increases in
the price of petroleum, though this situation would not be in direct
response to the CAFE/GHG standards. Thus, we cannot completely reject
the worst-case scenario for all vehicles, although we can and do
recognize that the footprint-based standards will significantly limit
the likelihood of its occurrence within the context of this rulemaking.
In summary, the agencies recognize the balancing inherent in
achieving higher levels of fuel economy and lower levels of
CO2 emissions through reduction of vehicle mass. Based on
the 2010 Kahane analysis that attempts to separate the effects of mass
reductions and footprint reductions, and to account better for the
possibility that mass reduction will be accomplished entirely through
methods that preserves structural strength and vehicle safety, the
agencies now believe that the likely deleterious safety effects of the
MYs 2012-2016 standards may be much lower than originally estimated.
They may be close to zero, or possibly beneficial if mass reduction is
carefully undertaken in the future and if the mass reduction in the
heavier LTVs is greater (in absolute terms) than in passenger cars. In
light of these findings, we believe that the balancing is reasonable.
5. How do the agencies plan to address this issue going forward?
NHTSA and EPA believe that it is important for the agencies to
conduct further study and research into the interaction of mass, size
and safety to assist future rulemakings. The agencies intend to begin
working collaboratively and to explore with DOE, CARB, and perhaps
other stakeholders an interagency/intergovernmental working group to
evaluate all aspects of mass, size and safety. It would also be the
goal of this team to coordinate government supported studies and
independent research, to the extent possible, to help ensure the work
is complementary to previous and ongoing research and to guide further
research in this area. DOE's EERE office has long funded extensive
research into component advanced vehicle materials and vehicle mass
reduction. Other agencies may have additional expertise that will be
helpful in establishing a coordinated work plan. The agencies are
interested in looking at the weight-safety relationship in a more
holistic (complete vehicle) way, and thanks to this CAFE rulemaking
NHTSA has begun to bring together parts of the agency--crashworthiness,
and crash avoidance rulemaking offices and the agency's Research &
Development office--in an interdisciplinary way to better leverage the
expertise of the agency. Extending this effort to other agencies will
help to ensure that all aspects of the weight-safety relationship are
considered completely and carefully with our future research. The
agencies also intend to carefully consider comments received in
response to the NPRM in developing plans for future studies and
research and to solicit input from stakeholders.
The agencies also plan to watch for safety effects as the U.S.
light-duty vehicle fleet evolves in response both to the CAFE/GHG
standards and to consumer preferences over the next several years.
Additionally, as new and

[[Page 25396]]

advanced materials and component smart designs are developed and
commercialized, and as manufacturers implement them in more vehicles,
it will be useful for the agencies to learn more about them and to try
to track these vehicles in the fleet to understand the relationship
between vehicle design and injury/fatality data. Specifically, the
agencies intend to follow up with study and research of the following:
First, NHTSA is in the process of contracting with an independent
institution to review the statistical methods that NHTSA and DRI have
used to analyze historical data related to mass, size and safety, and
to provide recommendation on whether the existing methods or other
methods should be used for future statistical analysis of historical
data. This study will include a consideration of potential near
multicollinearity in the historical data and how best to address it in
a regression analysis. This study is being initiated because, in
response to the NPRM, NHTSA received a number of comments related to
the methodology NHTSA used for the NPRM to determine the relationship
between mass and safety, as discussed in detail above.
Second, NHTSA and EPA, in consultation with DOE, intend to begin
updating the MYs 1991-1999 database on which the safety analyses in the
NPRM and final rule are based with newer vehicle data in the next
several months. This task will take at least a year to complete. This
study is being initiated in response to the NPRM comments related to
the use of data from MYs 1991-1999 in the NHTSA analysis, as discussed
in detail above.
Third, in order to assess if the design of recent model year
vehicles that incorporate various mass reduction methods affect the
relationships among vehicle mass, size and safety, NHTSA and EPA intend
to conduct collaborative statistical analysis, beginning in the next
several months. The agencies intend to work with DOE to identify
vehicles that are using material substitution and smart design. After
these vehicles are identified, the agencies intend to assess if there
are sufficient data for statistical analysis. If there are sufficient
data, statistical analysis would be conducted to compare the
relationship among mass, size and safety of these smart design vehicles
to vehicles of similar size and mass with more traditional designs.
This study is being initiated because, in response to the NPRM, NHTSA
received comments related to the use of data from MYs 1991-1999 in the
NHTSA analysis that did not include new designs that might change the
relationship among mass, size and safety, as discussed in detail above.
NHTSA may initiate a two-year study of the safety of the fleet
through an analysis of the trends in structural stiffness and whether
any trends identified impact occupant injury response in crashes.
Vehicle manufacturers may employ stiffer light weight materials to
limit occupant compartment intrusion while controlling for mass that
may expose the occupants to higher accelerations resulting in a greater
chance of injury in real-world crashes. This study would provide
information that would increase the understanding of the effects on
safety of newer vehicle designs.
In addition, NHTSA and EPA, possibly in collaboration with DOE, may
conduct a longer-term computer modeling-based design and analysis study
to help determine the maximum potential for mass reduction in the MYs
2017-2021 timeframe, through direct material substitution and smart
design while meeting safety regulations and guidelines, and maintaining
vehicle size and functionality. This study may build upon prior
research completed on vehicle mass reduction. This study would further
explore the comprehensive vehicle effects, including dissimilar
material joining technologies, manufacturer feasibility of both
supplier and OEM, tooling costs, and crash simulation and perhaps
eventual crash testing.

III. EPA Greenhouse Gas Vehicle Standards

A. Executive Overview of EPA Rule

1. Introduction
The Environmental Protection Agency (EPA) is establishing GHG
emissions standards for the largest sources of transportation GHGs--
light-duty vehicles, light-duty trucks, and medium-duty passenger
vehicles (hereafter light vehicles). These vehicle categories, which
include cars, sport utility vehicles, minivans, and pickup trucks used
for personal transportation, are responsible for almost 60% of all U.S.
transportation related emissions of the six gases discussed above
(Section I.A). This action represents the first-ever EPA rule to
regulate vehicle GHG emissions under the Clean Air Act (CAA) and will
establish standards for model years 2012-2016 and later light vehicles
sold in the United States.
EPA is adopting three separate standards. The first and most
important is a set of fleet-wide average carbon dioxide
(CO2) emission standards for cars and trucks. These
standards are CO2 emissions-footprint curves, where each
vehicle has a different CO2 emissions compliance target
depending on its footprint value. Vehicle CO2 emissions will
be measured over the EPA city and highway tests. The rule allows for
credits based on demonstrated improvements in vehicle air conditioner
systems, including both efficiency and refrigerant leakage improvement,
which are not captured by the EPA tests. The EPA projects that the
average light vehicle tailpipe CO2 level in model year 2011
will be 325 grams per mile while the average vehicle fleetwide average
CO2 emissions compliance level for the model year 2016
standard will be 250 grams per mile, an average reduction of 23 percent
from today's CO2 levels.
EPA is also finalizing standards that will cap tailpipe nitrous
oxide (N2O) and methane (CH4) emissions at 0.010
and 0.030 grams per mile, respectively. Even after adjusting for the
higher relative global warming potencies of these two compounds,
nitrous oxide and methane emissions represent less than one percent of
overall vehicle greenhouse gas emissions from new vehicles.
Accordingly, the goal of these two standards is to limit any potential
increases of tailpipe emissions of these compounds in the future but
not to force reductions relative to today's low levels.
This final rule responds to the Supreme Court's 2007 decision in
Massachusetts v. EPA \147\ which found that greenhouse gases fit within
the definition of air pollutant in the Clean Air Act. The Court held
that the Administrator must determine whether or not emissions from new
motor vehicles cause or contribute to air pollution which may
reasonably be anticipated to endanger public health or welfare, or
whether the science is too uncertain to make a reasoned decision. The
Court further ruled that, in making these decisions, the EPA
Administrator is required to follow the language of section 202(a) of
the CAA. The case was remanded back to the Agency for reconsideration
in light of the court's decision.
---------------------------------------------------------------------------

\147\ 549 U.S.C. 497 (2007). For further information on
Massachusetts v. EPA see the Endangerment and Cause or Contribute
Findings for Greenhouse Gases under Section 202(a) the Clean Air
Act, published in the Federal Register on December 15, 2009 (74 FR
66496). There is a comprehensive discussion of the litigation's
history, the Supreme Court's findings, and subsequent actions
undertaken by the Bush Administration and the EPA from 2007-2008 in
response to the Supreme Court remand. This information is also
available at: http://www.epa.gov/climatechange/endangerment.html.
---------------------------------------------------------------------------

The Administrator has responded to the remand by issuing two
findings under section 202(a) of the Clean Air

[[Page 25397]]

Act.\148\ First, the Administrator found that the science supports a
positive endangerment finding that the mix of six greenhouse gases
(carbon dioxide (CO2), methane (CH4), nitrous
oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons
(PFCs), and sulfur hexafluoride (SF6)) in the atmosphere
endangers the public health and welfare of current and future
generations. This is referred to as the endangerment finding. Second,
the Administrator found that the combined emissions of the same six
gases from new motor vehicles and new motor vehicle engines contribute
to the atmospheric concentrations of these key greenhouse gases and
hence to the threat of climate change. This is referred to as the cause
and contribute finding. Motor vehicles and new motor vehicle engines
emit carbon dioxide, methane, nitrous oxide, and hydrofluorocarbons.
EPA provides more details below on the legal and scientific bases for
this final rule.
---------------------------------------------------------------------------

\148\ See 74 FR 66496 (Dec. 15, 2009), ``Endangerment and Cause
or Contribute Findings for Greenhouse Gases Under Section 202(a) of
the Clean Air Act''.
---------------------------------------------------------------------------

As discussed in Section I, this GHG rule is part of a joint
National Program such that a large majority of the projected benefits
are achieved jointly with NHTSA's CAFE rule which is described in
detail in Section IV of this preamble. EPA projects total
CO2 equivalent emissions savings of approximately 960
million metric tons as a result of the rule, and oil savings of 1.8
billion barrels over the lifetimes of the MY 2012-2016 vehicles subject
to the rule. EPA projects that over the lifetimes of the MY 2012-2016
vehicles, the rule will cost $52 billion but will result in benefits of
$240 billion at a 3 percent discount rate, or $192 billion at a 7
percent discount rate (both values assume the average SCC value at 3%,
i.e., the $21/ton SCC value in 2010). Accordingly, these light vehicle
greenhouse gas emissions standards represent an important contribution
under the Clean Air Act toward meeting long-term greenhouse gas
emissions and import oil reduction goals, while providing important
economic benefits as well. The results of our analysis of 2012-2016 MY
vehicles, which we refer to as our ``model year analysis,'' are
summarized in Tables III.H.10-4 to III.H.10-7.
We have also looked beyond the lifetimes of 2012-2016 MY vehicles
at annual costs and benefits of the program for the 2012 through 2050
timeframe. We refer to this as our ``calendar year'' analysis (as
opposed to the costs and benefits mentioned above which we refer to as
our ``model year analysis''). In our calendar year analysis, the new
2016 MY standards are assumed to apply to all vehicles sold in model
years 2017 and later. The net present values of annual costs for the
2012 through 2050 timeframe are $346 billion for new vehicle technology
which will provide $1.5 billion in fuel savings, both values at a 3
percent discount rate. At a 7 percent discount rate over the same
period, the technology costs are estimated at $192 billion which will
provide $673 billion in fuel savings. The social benefits during the
2012 through 2050 timeframe are estimated at $454 billion and $305
billion at a 3 and 7 percent discount rate, respectively. Both of these
benefit estimates assume the average SCC value at 3% (i.e., the $21/ton
SCC value in 2010). The net benefits during this time period are then
$1.7 billion and $785 million at a 3 and 7 percent discount rate,
respectively. The results of our ``calendar year'' analysis are
summarized in Tables III.H 10-1 to III.H.10-3.
2. Why is EPA establishing this Rule?
This rule addresses only light vehicles. EPA is addressing light
vehicles as a first step in control of greenhouse gas emissions under
the Clean Air Act for four reasons. First, light vehicles are
responsible for almost 60% of all mobile source GHG emissions, a share
three times larger than any other mobile source subsector, and
represent about one-sixth of all U.S. greenhouse gas emissions. Second,
technology exists that can be readily and cost-effectively applied to
these vehicles to reduce their greenhouse gas emissions in the near
term. Third, EPA already has an existing testing and compliance program
for these vehicles, refined since the mid-1970s for emissions
compliance and fuel economy determinations, which would require only
minor modifications to accommodate greenhouse gas emissions
regulations. Finally, this rule is an important step in responding to
the Supreme Court's ruling in Massachusetts v. EPA, which applies to
other emissions sources in addition to light-duty vehicles. In fact,
EPA is currently evaluating controls for motor vehicles other than
those covered by this rule, and is also reviewing seven motor vehicle
related petitions submitted by various states and organizations
requesting that EPA use its Clean Air Act authorities to take action to
reduce greenhouse gas emissions from aircraft (under Sec. 231(a)(2)),
ocean-going vessels (under Sec. 213(a)(4)), and other nonroad engines
and vehicle sources (also under Sec. 213(a)(4)).
a. Light Vehicle Emissions Contribute to Greenhouse Gases and the
Threat of Climate Change
Greenhouse gases are gases in the atmosphere that effectively trap
some of the Earth's heat that would otherwise escape to space.
Greenhouse gases are both naturally occurring and anthropogenic. The
primary greenhouse gases of concern that are directly emitted by human
activities include carbon dioxide, methane, nitrous oxide,
hydrofluorocarbons, perfluorocarbons, and sulfur hexafluoride.
These gases, once emitted, remain in the atmosphere for decades to
centuries. Thus, they become well mixed globally in the atmosphere and
their concentrations accumulate when emissions exceed the rate at which
natural processes remove greenhouse gases from the atmosphere. The
heating effect caused by the human-induced buildup of greenhouse gases
in the atmosphere is very likely the cause of most of the observed
global warming over the last 50 years.\149\ The key effects of climate
change observed to date and projected to occur in the future include,
but are not limited to, more frequent and intense heat waves, more
severe wildfires, degraded air quality, heavier and more frequent
downpours and flooding, increased drought, greater sea level rise, more
intense storms, harm to water resources, continued ocean acidification,
harm to agriculture, and harm to wildlife and ecosystems. A detailed
explanation of observed and projected changes in greenhouse gases and
climate change and its impact on health, society, and the environment
is included in EPA's technical support document for the recently
promulgated Endangerment and Cause or Contribute Findings for
Greenhouse Gases Under Section 202(a) of the Clean Air Act.\150\
---------------------------------------------------------------------------

\149\ ``Technical Support Document for Endangerment and Cause or
Contribute Findings for Greenhouse Gases Under Section 202(a) of the
Clean Air Act'' Docket: EPA-HQ-OAR-2009-0472-11292.
\150\ 74 FR 66496 (Dec. 15, 2009). Both the Federal Register
Notice and the Technical Support Document for Endangerment and Cause
or Contribute Findings are found in the public docket No. EPA-OAR-
2009-0171, in the public docket established for this rulemaking, and
at http://epa.gov/climatechange/endangerment.html.
---------------------------------------------------------------------------

Mobile sources represent a large and growing share of United States
greenhouse gases and include light-duty vehicles, light-duty trucks,
medium-duty passenger vehicles, heavy duty trucks, airplanes,
railroads, marine vessels and a variety of other sources. In 2007, all
mobile sources emitted 31% of

[[Page 25398]]

all U.S. GHGs, and were the fastest-growing source of U.S. GHGs in the
U.S. since 1990. Transportation sources, which do not include certain
off-highway sources such as farm and construction equipment, account
for 28% of U.S. GHG emissions, and Section 202(a) sources, which
include light-duty vehicles, light-duty trucks, medium-duty passenger
vehicles, heavy-duty trucks, buses, and motorcycles account for 23% of
total U.S. GHGs.\151\
---------------------------------------------------------------------------

\151\ Inventory of U.S. Greenhouse Gases and Sinks: 1990-2007.
---------------------------------------------------------------------------

Light vehicles emit carbon dioxide, methane, nitrous oxide and
hydrofluorocarbons. Carbon dioxide (CO2) is the end product
of fossil fuel combustion. During combustion, the carbon stored in the
fuels is oxidized and emitted as CO2 and smaller amounts of
other carbon compounds.\152\ Methane (CH4) emissions are a
function of the methane content of the motor fuel, the amount of
hydrocarbons passing uncombusted through the engine, and any post-
combustion control of hydrocarbon emissions (such as catalytic
converters).\153\ Nitrous oxide (N2O) (and nitrogen oxide
(NOX)) emissions from vehicles and their engines are closely
related to air-fuel ratios, combustion temperatures, and the use of
pollution control equipment. For example, some types of catalytic
converters installed to reduce motor vehicle NOX, carbon
monoxide (CO) and hydrocarbon emissions can promote the formation of
N2O.\154\ Hydrofluorocarbons (HFC) emissions are
progressively replacing chlorofluorocarbons (CFC) and
hydrochlorofluorocarbons (HCFC) in these vehicles' cooling and
refrigeration systems as CFCs and HCFCs are being phased out under the
Montreal Protocol and Title VI of the CAA. There are multiple emissions
pathways for HFCs with emissions occurring during charging of cooling
and refrigeration systems, during operations, and during
decommissioning and disposal.\155\
---------------------------------------------------------------------------

\152\ Mobile source carbon dioxide emissions in 2006 equaled 26
percent of total U.S. CO2 emissions.
\153\ In 2006, methane emissions equaled 0.32 percent of total
U.S. methane emissions. Nitrous oxide is a product of the reaction
that occurs between nitrogen and oxygen during fuel combustion.
\154\ In 2006, nitrous oxide emissions for these sources
accounted for 8 percent of total U.S. nitrous oxide emissions.
\155\ In 2006, HFC from these source categories equaled 56
percent of total U.S. HFC emissions, making it the single largest
source category of U.S. HFC emissions.
---------------------------------------------------------------------------

b. Basis for Action Under the Clean Air Act
Section 202(a)(1) of the Clean Air Act (CAA) states that ``the
Administrator shall by regulation prescribe (and from time to time
revise) * * * standards applicable to the emission of any air pollutant
from any class or classes of new motor vehicles * * *, which in his
judgment cause, or contribute to, air pollution which may reasonably be
anticipated to endanger public health or welfare.'' As noted above, the
Administrator has found that the elevated concentrations of greenhouse
gases in the atmosphere may reasonably be anticipated to endanger
public health and welfare.\156\ The Administrator defined the ``air
pollution'' referred to in CAA section 202(a) to be the combined mix of
six long-lived and directly emitted GHGs: Carbon dioxide
(CO2), methane (CH4), nitrous oxide
(N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs),
and sulfur hexafluoride (SF6). The Administrator has further
found under CAA section 202(a) that emissions of the single air
pollutant defined as the aggregate group of these same six greenhouse
gases from new motor vehicles and new motor vehicle engines contribute
to air pollution. As a result of these findings, section 202(a)
requires EPA to issue standards applicable to emissions of that air
pollutant. New motor vehicles and engines emit CO2, methane,
N2O and HFC. This preamble describes the provisions that
control emissions of CO2, HFCs, nitrous oxide, and methane.
For further discussion of EPA's authority under section 202(a), see
Section I.C.2 of the preamble to the proposed rule (74 FR at 49464-66).
---------------------------------------------------------------------------

\156\ 74 FR 66496 (Dec. 15, 2009).
---------------------------------------------------------------------------

There are a variety of other CAA Title II provisions that are
relevant to standards established under section 202(a). The standards
are applicable to motor vehicles for their useful life. EPA has the
discretion in determining what standard applies over the vehicles'
useful life and has exercised that discretion in this rule. See Section
III.E.4 below.
The standards established under CAA section 202(a) are implemented
and enforced through various mechanisms. Manufacturers are required to
obtain an EPA certificate of conformity before they may sell or
introduce their new motor vehicle into commerce, according to CAA
section 206(a). The introduction into commerce of vehicles without a
certificate of conformity is a prohibited act under CAA section 203
that may subject a manufacturer to civil penalties and injunctive
actions (see CAA sections 204 and 205). Under CAA section 206(b), EPA
may conduct testing of new production vehicles to determine compliance
with the standards. For in-use vehicles, if EPA determines that a
substantial number of vehicles do not conform to the applicable
regulations then the manufacturer must submit and implement a remedial
plan to address the problem (see CAA section 207(c)). There are also
emissions-based warranties that the manufacturer must implement under
CAA section 207(a). Section III.E describes the rule's certification,
compliance, and enforcement mechanisms.
c. EPA's Endangerment and Cause or Contribute Findings for Greenhouse
Gases Under Section 202(a) of the Clean Air Act
On December 7, 2009 EPA's Administrator signed an action with two
distinct findings regarding greenhouse gases under section 202(a) of
the Clean Air Act. On December 15, 2009, the final findings were
published in the Federal Register. This action is called the
Endangerment and Cause or Contribute Findings for Greenhouse Gases
under Section 202(a) of the Clean Air Act (Endangerment Finding).\157\
Below are the two distinct findings:
---------------------------------------------------------------------------

\157\ 74 FR 66496 (Dec. 15, 2009)
---------------------------------------------------------------------------

Endangerment Finding: The Administrator finds that the
current and projected concentrations of the six key well-mixed
greenhouse gases--carbon dioxide (CO2), methane
(CH4), nitrous oxide (N2O), hydrofluorocarbons
(HFCs), perfluorocarbons (PFCs), and sulfur hexafluoride
(SF6)--in the atmosphere threaten the public health and
welfare of current and future generations.
Cause or Contribute Finding: The Administrator finds that
the combined emissions of these well-mixed greenhouse gases from new
motor vehicles and new motor vehicle engines contribute to the
greenhouse gas pollution which threatens public health and welfare.
Specifically, the Administrator found, after a thorough examination
of the scientific evidence on the causes and impact of current and
future climate change, and careful review of public comments, that the
science compellingly supports a positive finding that atmospheric
concentrations of these greenhouse gases result in air pollution which
may reasonably be anticipated to endanger both public health and
welfare. In her finding, the Administrator relied heavily upon the
major findings and conclusions from the

[[Page 25399]]

recent assessments of the U.S. Climate Change Science Program and the
U.N. Intergovernmental Panel on Climate Change.\158\ The Administrator
made a positive endangerment finding after considering both observed
and projected future effects of climate change, key uncertainties, and
the full range of risks and impacts to public health and welfare
occurring within the United States. In addition, the finding focused on
impacts within the U.S. but noted that the evidence concerning risks
and impacts occurring outside the U.S. provided further support for the
finding.
---------------------------------------------------------------------------

\158\ The U.S. Climate Change Science Program (CCSP) is now
called the U.S. Global Change Research Program (GCRP).
---------------------------------------------------------------------------

The key scientific findings supporting the endangerment finding are
that:

-- Concentrations of greenhouse gases are at unprecedented levels
compared to recent and distant past. These high concentrations are the
unambiguous result of anthropogenic emissions and are very likely the
cause of the observed increase in average temperatures and other
climatic changes.
-- The effects of climate change observed to date and projected to
occur in the future include more frequent and intense heat waves, more
severe wildfires, degraded air quality, heavier downpours and flooding,
increasing drought, greater sea level rise, more intense storms, harm
to water resources, harm to agriculture, and harm to wildlife and
ecosystems. These impacts are effects on public health and welfare
within the meaning of the Clean Air Act.

The Administrator found that emissions of the single air pollutant
defined as the aggregate group of these same six greenhouse gases from
new motor vehicles and new motor vehicle engines contribute to the air
pollution and hence to the threat of climate change. Key facts
supporting this cause and contribute finding for on-highway vehicles
regulated under section 202(a) of the Clean Air Act are that these
sources are responsible for 24% of total U.S. greenhouse gas emissions,
and more than 4% of total global greenhouse gas emissions.\159\ As
noted above, these findings require EPA to issue standards under
section 202(a) ``applicable to emission'' of the air pollutant that EPA
found causes or contributes to the air pollution that endangers public
health and welfare. The final emissions standards satisfy this
requirement for greenhouse gases from light-duty vehicles. Under
section 202(a) the Administrator has significant discretion in how to
structure the standards that apply to the emission of the air pollutant
at issue here, the aggregate group of six greenhouse gases. EPA has the
discretion under section 202(a) to adopt separate standards for each
gas, a single composite standard covering various gases, or any
combination of these. In this rulemaking EPA is finalizing separate
standards for nitrous oxide and methane, and a CO2 standard
that provides for credits based on reductions of HFCs, as the
appropriate way to issue standards applicable to emission of the single
air pollutant, the aggregate group of six greenhouse gases. EPA is not
setting any standards for perfluorocarbons (PFCs) or sulfur
hexafluoride (SF6) as they are not emitted by motor
vehicles.
---------------------------------------------------------------------------

\159\ This figure includes the greenhouse gas contributions of
light vehicles, heavy duty vehicles, and remaining on-highway mobile
sources. Light-duty vehicles are responsible for over 70 percent of
Section 202(a) mobile source GHGs, or about 17% of total U.S.
greenhouse gas emissions. U.S. EPA.2009 Technical Support Document
for Endangerment and Cause or Contribute Findings for Greenhouse
Gases under Section 202(a) of the Clean Air Act. Washington, DC. pp.
180-194. Available at http://epa.gov/climatechange/endangerment/
downloads/Endangerment%20TSD.pdf.
---------------------------------------------------------------------------

3. What is EPA adopting?
a. Light-Duty Vehicle, Light-Duty Truck, and Medium-Duty Passenger
Vehicle Greenhouse Gas Emission Standards and Projected Compliance
Levels
The following section provides an overview of EPA's final rule. The
key public comments are not discussed here, but are discussed in the
sections that follow which provide the details of the program. Comments
are also discussed in the Response to Comments document.
The CO2 emissions standards are by far the most
important of the three standards and are the primary focus of this
summary. As proposed, EPA is adopting an attribute-based approach for
the CO2 fleet-wide standard (one for cars and one for
trucks), using vehicle footprint as the attribute. These curves
establish different CO2 emissions targets for each unique
car and truck footprint. Generally, the larger the vehicle footprint,
the higher the corresponding vehicle CO2 emissions target.
Table III.A.3-1 shows the greenhouse gas standards for light vehicles
that EPA is finalizing for model years (MY) 2012 and later:

Table III.A.3-1--Industry-Wide Greenhouse Gas Emissions Standards
----------------------------------------------------------------------------------------------------------------
Standard/covered compounds Form of standard Level of standard Credits Test cycles
----------------------------------------------------------------------------------------------------------------
CO2 Standard: \160\ Tailpipe CO2 Fleetwide average Projected CO2-e credits\161\ EPA 2-cycle (FTP
footprint CO2- Fleetwide CO2 and HFET test
curves for cars level of 250 g/mi cycles).\162\
and trucks. (See footprint
curves in Sec.
III.B.2).
N2O Standard: Tailpipe N2O...... Cap per vehicle... 0.010 g/mi........ None *............ EPA FTP test.
CH4 Standard: Tailpipe CH4...... Cap per vehicle... 0.030 g/mi........ None *............ EPA FTP test.
----------------------------------------------------------------------------------------------------------------
* For N2O and CH4, manufacturers may optionally demonstrate compliance with a CO2-equivalent standard equal to
its footprint-based CO2 target level, using the FTP and HFET tests.

One important flexibility associated with the CO2
standard is the option for

[[Page 25400]]

manufacturers to obtain credits associated with improvements in their
air conditioning systems. EPA is adopting the air conditioning
provisions with minor modifications. As will be discussed in greater
detail in later sections, EPA is establishing test procedures and
design criteria by which manufacturers can demonstrate improvements in
both air conditioner efficiency (which reduces vehicle tailpipe
CO2 by reducing the load on the engine) and air conditioner
refrigerants (using lower global warming potency refrigerants and/or
improving system design to reduce GHG emissions associated with leaks).
Neither of these strategies to reduce GHG emissions from air
conditioners will be reflected in the EPA FTP or HFET tests. These
improvements will be translated to a g/mi CO2-equivalent
credit that can be subtracted from the manufacturer's tailpipe
CO2 compliance value. EPA expects a high percentage of
manufacturers to use this flexibility to earn air conditioning-related
credits for MY 2012-2016 vehicles such that the average credit earned
is about 11 grams per mile CO2-equivalent in 2016.
---------------------------------------------------------------------------

\160\ While over 99 percent of the carbon in automotive fuels is
converted to CO2 in a properly functioning engine,
compliance with the CO2 standard will also account for
the very small levels of carbon associated with vehicle tailpipe
hydrocarbon (HC) and carbon monoxide (CO) emissions, converted to
CO2 on a mass basis, as discussed further in Section
III.B.
\161\ CO2-e refers to CO2-equivalent, and
is a metric that allows non-CO2 greenhouse gases (such as
hydrofluorocarbons used as automotive air conditioning refrigerants)
to be expressed as an equivalent mass (i.e., corrected for relative
global warming potency) of CO2 emissions.
\162\ FTP is the Federal Test Procedure which uses what is
commonly referred to as the ``city'' driving schedule, and HFET is
the Highway Fuel Economy Test which uses the ``highway'' driving
schedule. Compliance with the CO2 standard will be based
on the same 2-cycle values that are currently used for CAFE
standards compliance; EPA projects that fleet-wide in-use or real
world CO2 emissions are approximately 25 percent higher,
on average, than 2-cycle CO2 values. Separate mechanisms
apply for A/C credits.
---------------------------------------------------------------------------

A second flexibility, being finalized essentially as proposed, is
CO2 credits for flexible and dual fuel vehicles, similar to
the CAFE credits for such vehicles which allow manufacturers to gain up
to 1.2 mpg in their overall CAFE ratings. The Energy Independence and
Security Act of 2007 (EISA) mandated a phase-out of these flexible fuel
vehicle CAFE credits beginning in 2015, and ending after 2019. EPA is
allowing comparable CO2 credits for flexible fuel vehicles
through MY 2015, but for MY 2016 and beyond, the GHG rule treats
flexible and dual fuel vehicles on a CO2-performance basis,
calculating the overall CO2 emissions for flexible and dual
fuel vehicles based on a fuel use-weighted average of the
CO2 levels on gasoline and on the alternative fuel, and on a
manufacturer's demonstration of actual usage of the alternative fuel in
its vehicle fleet.
Table III.A.3-2 summarizes EPA projections of industry-wide 2-cycle
CO2 emissions and fuel economy levels that will be achieved
by manufacturer compliance with the GHG standards for MY 2012-2016.
For MY 2011, Table III.A.3-2 uses the NHTSA projections of the
average fuel economy level that will be achieved by the MY 2011 fleet
of 30.8 mpg for cars and 23.3 mpg for trucks, converted to an
equivalent combined car and truck CO2 level of 326 grams per
mile.\163\ EPA believes this is a reasonable estimate with which to
compare the MY 2012-2016 CO2 emission standards. Identifying
the proper MY 2011 estimate is complicated for many reasons, among them
being the turmoil in the current automotive market for consumers and
manufacturers, uncertain and volatile oil and gasoline prices, the
ability of manufacturers to use flexible fuel vehicle credits to meet
MY 2011 CAFE standards, and the fact that most manufacturers have been
surpassing CAFE standards (particularly the car standard) in recent
years. Taking all of these considerations into account, EPA believes
that the MY 2011 projected CAFE achieved values, converted to
CO2 emissions levels, represent a reasonable estimate.
---------------------------------------------------------------------------

\163\ As discussed in Section IV of this preamble.
---------------------------------------------------------------------------

Table III.A.3-2 shows projected industry-wide average
CO2 emissions values. The Projected CO2 Emissions
for the Footprint-Based Standard column shows the CO2 g/mi
level corresponding with the footprint standard that must be met. It is
based on the promulgated CO2-footprint curves and projected
footprint values, and will decrease each year to 250 grams per mile (g/
mi) in MY 2016. For MY 2012-2016, the emissions impact of the projected
utilization of flexible fuel vehicle (FFV) credits and the temporary
lead-time allowance alternative standard (TLAAS, discussed below) are
shown in the next two columns. The Projected CO2 Emissions
column gives the CO2 emissions levels projected to be
achieved given use of the flexible fuel credits and temporary lead-time
allowance program. This column shows that, relative to the MY 2011
estimate, EPA projects that MY 2016 CO2 emissions will be
reduced by 23 percent over five years. The Projected A/C Credit column
represents the industry wide average air conditioner credit
manufacturers are expected to earn on an equivalent CO2 gram
per mile basis in a given model year. In MY 2016, the projected A/C
credit of 10.6 g/mi represents 14 percent of the 76 g/mi CO2
emissions reductions associated with the final standards. The Projected
2-cycle CO2 Emissions column shows the projected
CO2 emissions as measured over the EPA 2-cycle tests, which
will allow compliance with the standard assuming projected utilization
of the FFV, TLAAS, and A/C credits.

Table III.A.3-2--Projected Fleetwide CO2 Emissions Values
[Grams per mile]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Projected CO2
emissions for Projected 2-
Model year the footprint- Projected FFV Projected Projected CO2 Projected A/C cycle CO2
based credit TLAAS credit emissions credit emissions
standard
--------------------------------------------------------------------------------------------------------------------------------------------------------
2011................................................... .............. .............. .............. (326) .............. (326)
2012................................................... 295 6.5 1.2 303 3.5 307
2013................................................... 286 5.8 0.9 293 5.0 298
2014................................................... 276 5.0 0.6 282 7.5 290
2015................................................... 263 3.7 0.3 267 10.0 277
2016................................................... 250 0.0 0.1 250 10.6 261
--------------------------------------------------------------------------------------------------------------------------------------------------------

EPA is also finalizing a series of flexibilities for compliance
with the CO2 standard which are not expected to
significantly affect the projected compliance and achieved values shown
above, but which should reduce the costs of achieving those reductions.
These flexibilities include the ability to earn: Annual credits for a
manufacturer's over-compliance with its unique fleet-wide average
standard, early credits from MY 2009-2011, credit for ``off-cycle''
CO2 reductions from new and innovative technologies that are
not reflected in CO2/fuel economy tests, as

[[Page 25401]]

well as the carry-forward and carry-backward of credits, and the
ability to transfer credits between a manufacturer's car and truck
fleets. These flexibilities are being adopted with only very minor
changes from the proposal, as discussed in Section III.C.
EPA is finalizing an incentive to encourage the commercialization
of advanced GHG/fuel economy control technologies, including electric
vehicles (EVs), plug-in hybrid electric vehicles (PHEVs), and fuel cell
vehicles (FCVs), for model years 2012-2016. EPA's proposal included an
emissions compliance value of zero grams/mile for EVs and FCVs, and the
electric portion of PHEVs, and a multiplier in the range of 1.2 to 2.0,
so that each advanced technology vehicle would count as greater than
one vehicle in a manufacturer's fleet-wide compliance calculation.
Several commenters were very concerned about these credits and upon
considering the public comments on this issue, EPA is finalizing an
advanced technology vehicle incentive program to assign a zero gram/
mile emissions compliance value for EVs and FCVs, and the electric
portion of PHEVs, for up to the first 200,000 EV/PHEV/FCV vehicles
produced by a given manufacturer during MY 2012-2016. For any
production greater than this amount, the compliance value for the
vehicle will be greater than zero gram/mile, set at a level that
reflects the vehicle's average net increase in upstream greenhouse gas
emissions in comparison to the gasoline or diesel vehicle it replaces.
EPA is not finalizing a multiplier based on the concerns potentially
excessive credits using that incentive. EPA agrees that the multiplier,
in combination with the zero grams/mile compliance value, would be
excessive. EPA will also allow this early advanced technology incentive
program beginning in MYs 2009 through 2011. Further discussion on the
advanced technology vehicle incentives, including more detail on the
public comments and EPA's response, is found in Section III.C.
EPA is also finalizing a temporary lead-time allowance (TLAAS) for
manufacturers that sell vehicles in the U.S. in MY 2009 and for which
U.S. vehicle sales in that model year are below 400,000 vehicles. This
allowance will be available only during the MY 2012-2015 phase-in years
of the program. A manufacturer that satisfies the threshold criteria
will be able to treat a limited number of vehicles as a separate
averaging fleet, which will be subject to a less stringent GHG
standard.\164\ Specifically, a standard of 125 percent of the vehicle's
otherwise applicable foot-print target level will apply to up to
100,000 vehicles total, spread over the four-year period of MY 2012
through 2015. Thus, the number of vehicles to which the flexibility
could apply is limited. EPA also is setting appropriate restrictions on
credit use for these vehicles, as discussed further in Section III. By
MY 2016, these allowance vehicles must be averaged into the
manufacturer's full fleet (i.e., they will no longer be eligible for a
different standard). EPA discusses this in more detail in Section III.B
of the preamble.
---------------------------------------------------------------------------

\164\ EPCA does not permit such an allowance. Consequently,
manufacturers who may be able to take advantage of a lead-time
allowance under the GHG standards would be required to comply with
the applicable CAFE standard or be subject to penalties for non-
compliance.
---------------------------------------------------------------------------

EPA received comments from several smaller manufacturers that the
TLAAS program was insufficient to allow manufacturers with very limited
product lines to comply. These manufacturers commented that they need
additional lead-time to meet the standards, because their
CO2 baselines are significantly higher and their vehicle
product lines are even more limited, reducing their ability to average
across their fleets compared even to other TLAAS manufacturers. EPA
fully summarizes the public comments on the TLAAS program, including
comments not supporting the program, in Section III.B. In summary, in
response to the lead time issues raised by manufacturers, EPA is
modifying the TLAAS program that applies to manufacturers with between
5,000 and 50,000 U.S. vehicle sales in MY 2009. These manufactures
would have an increased allotment of vehicles, a total of 250,000,
compared to 100,000 vehicles for other TLAAS-eligible manufacturers. In
addition, the TLAAS program for these manufacturers would be extended
by one year, through MY 2016 for these vehicles, for a total of five
years of eligibility. The other provisions of the TLAAS program would
continue to apply, such as the restrictions on credit trading and the
level of the standard. Additional restrictions would also apply to
these vehicles, as discussed in Section III.B.5. In addition, for the
smallest volume manufacturers, those with U.S. sales of below 5,000
vehicles, EPA is not setting standards at this time but is instead
deferring standards until a future rulemaking. This is the same
approach we are using for small businesses. The unique issues involved
with these manufacturers will be addressed in that future rulemaking.
Further discussion of the public comment on these issues and details on
these changes from the proposed program are included in Section
III.B.6. The agency received comments on its compliance with the
Regulatory Flexibility Act. As stated in Section III.I.3, small
entities are not significantly impacted by this rulemaking.
EPA is also adopting caps on the tailpipe emissions of nitrous
oxide (N2O) and methane (CH4)--0.010 g/mi for
N2O and 0.030 g/mi for CH4--over the EPA FTP
test. While N2O and CH4 can be potent greenhouse
gases on a relative mass basis, their emission levels from modern
vehicle designs are extremely low and represent only about 1% of total
late model light vehicle GHG emissions. These cap standards are
designed to ensure that N2O and CH4 emissions
levels do not rise in the future, rather than to force reductions in
the already low emissions levels. Accordingly, these standards are not
designed to require automakers to make any changes in current vehicle
designs, and thus EPA is not projecting any environmental or economic
costs or benefits associated with these standards.
EPA has attempted to build on existing practice wherever possible
in designing a compliance program for the GHG standards. In particular,
the program structure will streamline the compliance process for both
manufacturers and EPA by enabling manufacturers to use a single data
set to satisfy both the new GHG and CAFE testing and reporting
requirements. Timing of certification, model-level testing, and other
compliance activities also follow current practices established under
the Tier 2 emissions and CAFE programs.
EPA received numerous comments on issues related to the impacts on
stationary sources, due to the Clean Air Act's provisions for
permitting requirements related to the issuance of the proposed GHG
standards for new motor vehicles. Some comments suggested that EPA had
underestimated the number of stationary sources that may be subject to
GHG permitting requirements; other comments suggested that EPA did not
adequately consider the permitting impact on small business sources.
Other comments related to EPA's interpretation of the CAA's provisions
for subjecting stationary sources to permit regulation after GHG
standards are set. EPA's response to these comments is contained in the
Response to Comments document; however, many of these comments pertain
to issues that EPA is addressing in its consideration of the final
Greenhouse Gas Permit Tailoring

[[Page 25402]]

Rule, Prevention of Significant Deterioration and Title V Greenhouse
Gas Tailoring Rule; Proposed Rule, 74 FR 55292 (October 27, 2009) and
will thus be fully addressed in that rulemaking.
Some of the comments relating to the stationary source permitting
issues suggested that EPA should defer setting GHG standards for new
motor vehicles to avoid such stationary source permitting impacts. EPA
is issuing these final GHG standards for light-duty vehicles as part of
its efforts to expeditiously respond to the Supreme Court's nearly
three year old ruling in Massachusetts v. EPA, 549 U.S. 497 (2007). In
that case, the Court held that greenhouse gases fit within the
definition of air pollutant in the Clean Air Act, and that EPA is
therefore compelled to respond to the rulemaking petition under section
202(a) by determining whether or not emissions from new motor vehicles
cause or contribute to air pollution which may reasonably be
anticipated to endanger public health or welfare, or whether the
science is too uncertain to make a reasoned decision. The Court further
ruled that, in making these decisions, the EPA Administrator is
required to follow the language of section 202(a) of the CAA. The Court
stated that under section 202(a), ``[i]f EPA makes [the endangerment
and cause or contribute findings], the Clean Air Act requires the
agency to regulate emissions of the deleterious pollutant.'' 549 U.S.
at 534. As discussed above, EPA has made the two findings on
contribution and endangerment. 74 FR 66496 (December 15, 2009). Thus,
EPA is required to issue standards applicable to emissions of this air
pollutant from new motor vehicles.
The Court properly noted that EPA retained ``significant latitude''
as to the ``timing * * * and coordination of its regulations with those
of other agencies'' (id.). However it has now been nearly three years
since the Court issued its opinion, and the time for delay has passed.
In the absence of these final standards, there would be three separate
Federal and State regimes independently regulating light-duty vehicles
to increase fuel economy and reduce GHG emissions: NHTSA's CAFE
standards, EPA's GHG standards, and the GHG standards applicable in
California and other states adopting the California standards. This
joint EPA-NHTSA program will allow automakers to meet all of these
requirements with a single national fleet because California has
indicated that it will accept compliance with EPA's GHG standards as
compliance with California's GHG standards. 74 FR at 49460. California
has not indicated that it would accept NHTSA's CAFE standards by
themselves. Without EPA's vehicle GHG standards, the states will not
offer the Federal program as an alternative compliance option to
automakers and the benefits of a harmonized national program will be
lost. California and several other states have expressed strong concern
that, without comparable Federal vehicle GHG standards, the states will
not offer the Federal program as an alternative compliance option to
automakers. Letter dated February 23, 2010 from Commissioners of
California, Maine, New Mexico, Oregon and Washington to Senators Harry
Reid and Mitch McConnell (Docket EPA-HQ-OAR-2009-0472-11400). The
automobile industry also strongly supports issuance of these rules to
allow implementation of the national program and avoid ``a myriad of
problems for the auto industry in terms of product planning, vehicle
distribution, adverse economic impacts and, most importantly, adverse
consequences for their dealers and customers.'' Letter dated March 17,
2010 from Alliance of Automobile Manufacturers to Senators Harry Reid
and Mitch McConnell, and Representatives Nancy Pelosi and John Boehner
(Docket EPA-HQ-OAR-2009-0472-11368). Thus, without EPA's GHG standards
as part of a Federal harmonized program, important GHG reductions as
well as benefits to the automakers and to consumers would be lost.\165\
In addition, delaying the rule would impose significant burdens and
uncertainty on automakers, who are already well into planning for
production of MY 2012 vehicles, relying on the ability to produce a
single national fleet. Delaying the issuance of this final rule would
very seriously disrupt the industry's plans.
---------------------------------------------------------------------------

\165\ As discussed elsewhere, EPA's GHG standards achieve
greater overall reductions in GHGs than NHTSA's CAFE standards.
---------------------------------------------------------------------------

Instead of delaying the LDV rule and losing the benefits of this
rule and the harmonized national program, EPA is directly addressing
concerns about stationary source permitting in other actions that EPA
is taking with regard to such permitting. That is the proper approach
to address the issue of stationary source permitting, as compared to
delaying the issuance of this rule for some undefined, indefinite time
period.
Some parties have argued that EPA's issuance of this light-duty
vehicle rule amounts to a denial of various administrative requests
pending before EPA, in which parties have requested that EPA reconsider
and stay the GHG endangerment finding published on December 15, 2009.
That is not an accurate characterization of the impact of this final
rule. EPA has not taken final action on these administrative requests,
and issuance of this vehicle rule is not final agency action,
explicitly or implicitly, on those requests. Currently, while we
carefully consider the pending requests for reconsideration on
endangerment, these final findings on endangerment and contribution
remain in place. Thus under section 202(a) EPA is obligated to
promulgate GHG motor vehicle standards, although there is no statutory
deadline for issuance of the light-duty vehicle rule or other motor
vehicle rules. In that context, issuance of this final light-duty
vehicle rule does no more than recognize the current status of the
findings--they are final and impose a rulemaking obligation on EPA,
unless and until we change them. In issuing the vehicle rule we are not
making a decision on requests to reconsider or stay the endangerment
finding, and are not in any way prejudicing or limiting EPA's
discretion in making a final decision on these administrative requests.
For discussion of comments on impacts on small entities and EPA's
compliance with the Regulatory Flexibility Act, see the discussion in
Section III.I.3.
b. Environmental and Economic Benefits and Costs of EPA's Standards
In Table III.A.3-3 EPA presents estimated annual net benefits for
the indicated calendar years. The table also shows the net present
values of those benefits for the calendar years 2012-2050 using both a
3 percent and a 7 percent discount rate. As discussed previously, EPA
recognizes that much of these same costs and benefits are also
attributable to the CAFE standard contained in this joint final rule.

[[Page 25403]]

Table III.A.3-3--Projected Quantifiable Benefits and Costs for CO2 Standard
[In million 2007$]
--------------------------------------------------------------------------------------------------------------------------------------------------------
2020 2030 2040 2050 NPV, 3% \a\ NPV, 7% \a\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Quantified Annual Costs\b\.............................. -$20,100 -$64,000 -$101,900 -$152,200 -$1,199,700 -$480,700
--------------------------------------------------------------------------------------------------------------------------------------------------------
Benefits From Reduced CO2 Emissions at Each Assumed SCC Value c d e
--------------------------------------------------------------------------------------------------------------------------------------------------------
Avg SCC at 5%........................................... 900 2,700 4,600 7,200 34,500 34,500
Avg SCC at 3%........................................... 3,700 8,900 14,000 21,000 176,700 176,700
Avg SCC at 2.5%......................................... 5,800 14,000 21,000 30,000 299,600 299,600
95th percentile SCC at 3%............................... 11,000 27,000 43,000 62,000 538,500 538,500
--------------------------------------------------------------------------------------------------------------------------------------------------------
Other Impacts
--------------------------------------------------------------------------------------------------------------------------------------------------------
Criteria Pollutant Benefits f g h i..................... B 1,200-1,300 1,200-1,300 1,200-1,300 21,000 14,000
Energy Security Impacts (price shock)................... 2,200 4,500 6,000 7,600 81,900 36,900
Reduced Refueling....................................... 2,400 4,800 6,300 8,000 87,900 40,100
Value of Increased Driving \j\.......................... 4,200 8,800 13,000 18,400 171,500 75,500
Accidents, Noise, Congestion............................ -2,300 -4,600 -6,100 -7,800 -84,800 -38,600
--------------------------------------------------------------------------------------------------------------------------------------------------------
Quantified Net Benefits at Each Assumed SCC Value c d e
--------------------------------------------------------------------------------------------------------------------------------------------------------
Avg SCC at 5%........................................... 27,500 81,500 127,000 186,900 1,511,700 643,100
Avg SCC at 3%........................................... 30,300 87,700 136,400 200,700 1,653,900 785,300
Avg SCC at 2.5%......................................... 32,400 92,800 143,400 209,700 1,776,800 908,200
95th percentile SCC at 3%............................... 37,600 105,800 165,400 241,700 2,015,700 1,147,100
--------------------------------------------------------------------------------------------------------------------------------------------------------
\a\ Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same discount rate used to discount the
value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer to
Section III.F for more detail.
\b\ Quantified annual costs are negative because of fuel savings (see Table III.H.10-1 for a breakdown of the vehicle technology costs and fuel
savings). The fuel savings outweigh the vehicle technology costs and, therefore, the costs are presented here are negative values.
\c\ Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected under this final rule. Although EPA has
not monetized the benefits of reductions in these non-CO2 emissions, the value of these reductions should not be interpreted as zero. Rather, the
reductions in non-CO2 GHGs will contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC Technical Support Document (TSD)
notes the difference between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to value non-CO2
emissions in future analyses.
\d\ Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates range as follows: for Average SCC at
5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at 2.5%: $35-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also
presents these SCC estimates.
\e\ Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same discount rate used to discount the
value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer to SCC
TSD for more detail.
\f\ Note that ``B'' indicates unquantified criteria pollutant benefits in the year 2020. For the final rule, we only modeled the rule's PM2.5- and ozone-
related impacts in the calendar year 2030. For the purposes of estimating a stream of future-year criteria pollutant benefits, we assume that the
benefits out to 2050 are equal to, and no less than, those modeled in 2030 as reflected by the stream of estimated future emission reductions. The NPV
of criteria pollutant-related benefits should therefore be considered a conservative estimate of the potential benefits associated with the final
rule.
\g\ The benefits presented in this table include an estimate of PM-related premature mortality derived from Laden et al., 2006, and the ozone-related
premature mortality estimate derived from Bell et al., 2004. If the benefit estimates were based on the ACS study of PM-related premature mortality
(Pope et al., 2002) and the Levy et al., 2005 study of ozone-related premature mortality, the values would be as much as 70% smaller.
\h\ The calendar year benefits presented in this table assume either a 3% discount rate in the valuation of PM-related premature mortality ($1,300
million) or a 7% discount rate ($1,200 million) to account for a twenty-year segmented cessation lag. Note that the benefits estimated using a 3%
discount rate were used to calculate the NPV using a 3% discount rate and the benefits estimated using a 7% discount rate were used to calculate the
NPV using a 7% discount rate. For benefits totals presented at each calendar year, we used the mid-point of the criteria pollutant benefits range
($1,250).
\i\ Note that the co-pollutant impacts presented here do not include the full complement of endpoints that, if quantified and monetized, would change
the total monetized estimate of impacts. The full complement of human health and welfare effects associated with PM and ozone remain unquantified
because of current limitations in methods or available data. We have not quantified a number of known or suspected health effects linked with ozone
and PM for which appropriate health impact functions are not available or which do not provide easily interpretable outcomes (e.g., changes in heart
rate variability). Additionally, we are unable to quantify a number of known welfare effects, including reduced acid and particulate deposition damage
to cultural monuments and other materials, and environmental benefits due to reductions of impacts of eutrophication in coastal areas.
\j\ Calculated using pre-tax fuel prices.

4. Basis for the GHG Standards Under Section 202(a)
EPA statutory authority under section 202(a)(1) of the Clean Air
Act (CAA) is discussed in more detail in Section I.C.2 of the proposed
rule (74 FR at 49464-65). The following is a summary of the basis for
the final GHG standards under section 202(a), which is discussed in
more detail in the following portions of Section III.
With respect to CO2 and HFCs, EPA is adopting attribute-
based light-duty car and truck standards that achieve large and
important emissions reductions of GHGs. EPA has evaluated the
technological feasibility of the standards, and the information and
analysis performed by EPA indicates that these standards are feasible
in the lead time provided. EPA and NHTSA have carefully evaluated the
effectiveness of individual technologies as well as the interactions
when technologies are combined. EPA's projection of the technology that
would be used to comply with the standards indicates that manufacturers
will be able to meet the standards by employing

[[Page 25404]]

a wide variety of technologies that are already commercially available
and can be incorporated into their vehicles at the time of redesign. In
addition to the consideration of the manufacturers' redesign cycle,
EPA's analysis also takes into account certain flexibilities that will
facilitate compliance especially in the early years of the program when
potential lead time constraints are most challenging. These
flexibilities include averaging, banking, and trading of various types
of credits. For the industry as a whole, EPA's projections indicate
that the standards can be met using technology that will be available
in the lead-time provided. At the same time, it must be noted that
because technology is commercially available today does not mean it can
automatically be incorporated fleet-wide during the model years in
question. As discussed below, and in detail in Section III.D.7, EPA and
NHTSA carefully analyzed issues of adequacy of lead time in determining
the level of the standards, and the agencies are convinced both that
lead time is sufficient to meet the standards but that major further
additions of technology across the fleet is not possible during these
model years.
To account for additional lead-time concerns for various
manufacturers of typically higher performance vehicles, EPA is adopting
a Temporary Lead-time Allowance similar to that proposed that will
further facilitate compliance for limited volumes of such vehicles in
the program's initial years. For a few very small volume manufacturers,
EPA is deferring standards pending later rulemaking.
EPA has also carefully considered the cost to manufacturers of
meeting the standards, estimating piece costs for all candidate
technologies, direct manufacturing costs, cost markups to account for
manufacturers' indirect costs, and manufacturer cost reductions
attributable to learning. In estimating manufacturer costs, EPA took
into account manufacturers' own practices such as making major changes
to model technology packages during a planned redesign cycle. EPA then
projected the average cost across the industry to employ this
technology, as well as manufacturer-by-manufacturer costs. EPA
considers the per vehicle costs estimated from this analysis to be
within a reasonable range in light of the emissions reductions and
benefits received. EPA projects, for example, that the fuel savings
over the life of the vehicles will more than offset the increase in
cost associated with the technology used to meet the standards.
EPA has also evaluated the impacts of these standards with respect
to reductions in GHGs and reductions in oil usage. For the lifetime of
the model year 2012-2016 vehicles we estimate GHG reductions of
approximately 960 million metric tons CO2 eq. and fuel
reductions of 1.8 billion barrels of oil. These are important and
significant reductions. EPA has also analyzed a variety of other
impacts of the standards, ranging from the standards' effects on
emissions of non-GHG pollutants, impacts on noise, energy, safety and
congestion. EPA has also quantified the cost and benefits of the
standards, to the extent practicable. Our analysis to date indicates
that the overall quantified benefits of the standards far outweigh the
projected costs. Utilizing a 3% discount rate, we estimate the total
net social benefits over the life of the model year 2012-2016 vehicles
is $192 billion, and the net present value of the net social benefits
of the standards through the year 2050 is $1.9 trillion dollars.\166\
These values are estimated at $136 billion and $787 billion,
respectively, using a 7% discount rate and the SCC discounted at 3
percent.\167\
---------------------------------------------------------------------------

\166\ Based on the mean SCC at 3 percent discount rate, which is
$21 per metric ton CO2 in 2010 rising to $45 per metric
ton CO2 in 2050.
\167\ SCC was discounted at 3 percent to maintain internal
consistency in the SCC calculations while all other benefits were
discounted at 7 percent. Specifically, the same discount rate used
to discount the value of damages from future CO2
emissions is used to calculate net present value of SCC.
---------------------------------------------------------------------------

Under section 202(a) EPA is called upon to set standards that
provide adequate lead-time for the development and application of
technology to meet the standards. EPA's standards satisfy this
requirement, as discussed above. In setting the standards, EPA is
called upon to weigh and balance various factors, and to exercise
judgment in setting standards that are a reasonable balance of the
relevant factors. In this case, EPA has considered many factors, such
as cost, impacts on emissions (both GHG and non-GHG), impacts on oil
conservation, impacts on noise, energy, safety, and other factors, and
has, where practicable, quantified the costs and benefits of the rule.
In summary, given the technical feasibility of the standard, the
moderate cost per vehicle in light of the savings in fuel costs over
the life time of the vehicle, the very significant reductions in
emissions and in oil usage, and the significantly greater quantified
benefits compared to quantified costs, EPA is confident that the
standards are an appropriate and reasonable balance of the factors to
consider under section 202(a). See Husqvarna AB v. EPA, 254 F. 3d 195,
200 (DC Cir. 2001) (great discretion to balance statutory factors in
considering level of technology-based standard, and statutory
requirement ``to [give appropriate] consideration to the cost of
applying * * * technology'' does not mandate a specific method of cost
analysis); see also Hercules Inc. v. EPA, 598 F. 2d 91, 106 (DC Cir.
1978) (``In reviewing a numerical standard we must ask whether the
agency's numbers are within a zone of reasonableness, not whether its
numbers are precisely right''); Permian Basin Area Rate Cases, 390 U.S.
747, 797 (1968) (same); Federal Power Commission v. Conway Corp., 426
U.S. 271, 278 (1976) (same); Exxon Mobil Gas Marketing Co. v. FERC, 297
F. 3d 1071, 1084 (DC Cir. 2002) (same).
EPA recognizes that the vast majority of technologies which we are
considering for purposes of setting standards under section 202(a) are
commercially available and already being utilized to a limited extent
across the fleet. The vast majority of the emission reductions, which
would result from this rule, would result from the increased use of
these technologies. EPA also recognizes that this rule would enhance
the development and limited use of more advanced technologies, such as
PHEVs and EVs. In this technological context, there is no clear cut
line that indicates that only one projection of technology penetration
could potentially be considered feasible for purposes of section
202(a), or only one standard that could potentially be considered a
reasonable balancing of the factors relevant under section 202(a). EPA
therefore evaluated two sets of alternative standards, one more
stringent than the promulgated standards and one less stringent.
The alternatives are 4% per year increase in standards which would
be less stringent and a 6% per year increase in the standards which
would be more stringent. EPA is not adopting either of these. As
discussed in Section III.D.7, the 4% per year forgoes CO2
reductions which can be achieved at reasonable cost and are achievable
by the industry within the rule's timeframe. The 6% per year
alternative requires a significant increase in the projected required
technology penetration which appears inappropriate in this timeframe
due to the limited available lead time and the current difficult
financial condition of the automotive industry. (See Section III.D.7
for a detailed discussion of why EPA is not adopting either of the
alternatives.) EPA also believes that the no backsliding standards it
is adopting

[[Page 25405]]

for N2O and CH4 are appropriate under section
202(a).

B. GHG Standards for Light-Duty Vehicles, Light-Duty Trucks, and
Medium-Duty Passenger Vehicles

EPA is finalizing new emission standards to control greenhouse
gases (GHGs) from light-duty vehicles. First, EPA is finalizing an
emission standard for carbon dioxide (CO2) on a gram per
mile (g/mile) basis that will apply to a manufacturer's fleet of cars,
and a separate standard that will apply to a manufacturer's fleet of
trucks. CO2 is the primary greenhouse gas resulting from the
combustion of vehicular fuels, and the amount of CO2 emitted
is directly correlated to the amount of fuel consumed. Second, EPA is
providing auto manufacturers with the opportunity to earn credits
toward the fleet-wide average CO2 standards for improvements
to air conditioning systems, including both hydrofluorocarbon (HFC)
refrigerant losses (i.e., system leakage) and indirect CO2
emissions related to the increased load on the engine. Third, EPA is
finalizing separate emissions standards for two other GHGs: Methane
(CH4) and nitrous oxide (N20). CH4 and
N2O emissions relate closely to the design and efficient use
of emission control hardware (i.e., catalytic converters). The
standards for CH4 and N2O will be set as a cap
that will limit emissions increases and prevent backsliding from
current emission levels. The final standards described below will apply
to passenger cars, light-duty trucks, and medium-duty passenger
vehicles (MDPVs). As an overall group, they are referred to in this
preamble as light vehicles or simply as vehicles. In this preamble
section passenger cars may be referred to simply as ``cars'', and
light-duty trucks and MDPVs as ``light trucks'' or ``trucks.'' \168\
---------------------------------------------------------------------------

\168\ As described in Section III.B.2., GHG emissions standards
will use the same vehicle category definitions as are used in the
CAFE program.
---------------------------------------------------------------------------

EPA's program includes a number of credit opportunities and other
flexibilities to help manufacturers comply, especially in the early
years of the program. EPA is establishing a system of averaging,
banking, and trading of credits integral to the fleet averaging
approach, based on manufacturer fleet average CO2
performance, as discussed in Section III.B.4. This approach is similar
to averaging, banking, and trading (ABT) programs EPA has established
in other programs and is also similar to provisions in the CAFE
program. In addition to traditional ABT credits based on the fleet
emissions average, EPA is also including A/C credits as an aspect of
the standards, as mentioned above. EPA is also including several
additional credit provisions that apply only in the initial model years
of the program. These include flex fuel vehicle credits, incentives for
the early commercialization of certain advanced technology vehicles,
credits for new and innovative ``off-cycle'' technologies that are not
captured by the current test procedures, and generation of credits
prior to model year 2012. The A/C credits and additional credit
opportunities are described in Section III.C. These credit programs
will provide flexibility to manufacturers, which may be especially
important during the early transition years of the program. EPA will
also allow a manufacturer to carry a credit deficit into the future for
a limited number of model years. A parallel provision, referred to as
credit carry-back, will be part of the CAFE program. Finally, EPA is
finalizing an optional compliance flexibility, the Temporary Leadtime
Allowance Alternative Standard program, for intermediate volume
manufacturers, and is deferring standards for the smallest
manufacturers, as discussed in Sections III.B.5 and 6 below.
1. What fleet-wide emissions levels correspond to the CO2
standards?
The attribute-based CO2 standards are projected to
achieve a national fleet-wide average, covering both light cars and
trucks, of 250 grams/mile of CO2 in model year (MY) 2016.
This includes CO2-equivalent emission reductions from A/C
improvements, reflected as credits in the standard. The standards will
begin with MY 2012, with a generally linear increase in stringency from
MY 2012 through MY 2016. EPA will have separate standards for cars and
light trucks. The tables in this section below provide overall fleet
average levels that are projected for both cars and light trucks over
the phase-in period which is estimated to correspond with the
standards. The actual fleet-wide average g/mi level that will be
achieved in any year for cars and trucks will depend on the actual
production for that year, as well as the use of the various credit and
averaging, banking, and trading provisions. For example, in any year,
manufacturers may generate credits from cars and use them for
compliance with the truck standard. Such transfer of credits between
cars and trucks is not reflected in the table below. In Section III.F,
EPA discusses the year-by-year estimate of emissions reductions that
are projected to be achieved by the standards.
In general, the schedule of standards acts as a phase-in to the MY
2016 standards, and reflects consideration of the appropriate lead-time
for each manufacturer to implement the requisite emission reductions
technology across its product line.\169\ Note that 2016 is the final
model year in which standards become more stringent. The 2016
CO2 standards will remain in place for 2017 and later model
years, until revised by EPA in a future rulemaking.
---------------------------------------------------------------------------

\169\ See CAA section 202(a)(2).
---------------------------------------------------------------------------

EPA estimates that, on a combined fleet-wide national basis, the
2016 MY standards will achieve a level of 250 g/mile CO2,
including CO2-equivalent credits from A/C related
reductions. The derivation of the 250 g/mile estimate is described in
Section III.B.2.
EPA has estimated the overall fleet-wide CO2-equivalent
emission levels that correspond with the attribute-based standards,
based on the projections of the composition of each manufacturer's
fleet in each year of the program. Tables III.B.1-1 and III.B.1-2
provides these estimates for each manufacturer.\170\
---------------------------------------------------------------------------

\170\ These levels do not include the effect of flexible fuel
credits, transfer of credits between cars and trucks, temporary lead
time allowance, or any other credits.
---------------------------------------------------------------------------

As a result of public comments and updated economic and future
fleet projections, the attribute based curves have been updated for
this final rule, as discussed in detail in Section II.B of this
preamble and Chapter 2 of the Joint TSD. This update in turn affects
costs, benefits, and other impacts of the final standards--thus EPA's
overall projection of the impacts of the final rule standards have been
updated and the results are different than for the NPRM, though in
general not by a large degree.

[[Page 25406]]

Table III.B.1-1--Estimated Fleet CO2-Equivalent Levels Corresponding to the Standards for Cars
[g/mile]
----------------------------------------------------------------------------------------------------------------
Model year
Manufacturer -------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
BMW............................. 266 259 250 239 228
Chrysler........................ 269 262 254 243 232
Daimler......................... 274 267 259 249 238
Ford............................ 267 259 251 240 229
General Motors.................. 268 261 252 241 230
Honda........................... 260 252 244 233 222
Hyundai......................... 260 254 246 233 222
Kia............................. 263 255 247 235 224
Mazda........................... 260 252 243 232 221
Mitsubishi...................... 257 249 241 230 219
Nissan.......................... 263 256 248 237 226
Porsche......................... 244 237 228 217 206
Subaru.......................... 253 246 237 226 215
Suzuki.......................... 245 238 230 218 208
Tata............................ 288 280 272 261 250
Toyota.......................... 259 251 243 232 221
Volkswagen...................... 256 249 240 229 219
----------------------------------------------------------------------------------------------------------------

Table III.B.1-2--Estimated Fleet CO2-Equivalent Levels Corresponding to the Standards for Light Trucks
[g/mile]
----------------------------------------------------------------------------------------------------------------
Model year
Manufacturer -------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
BMW............................. 330 320 310 297 283
Chrysler........................ 342 333 323 309 295
Daimler......................... 343 332 323 308 294
Ford............................ 354 344 334 319 305
General Motors.................. 364 354 344 330 316
Honda........................... 327 318 309 295 281
Hyundai......................... 325 316 307 292 278
Kia............................. 335 327 318 303 289
Mazda........................... 319 308 299 285 271
Mitsubishi...................... 316 306 297 283 269
Nissan.......................... 343 334 323 308 294
Porsche......................... 334 325 315 301 287
Subaru.......................... 315 305 296 281 267
Suzuki.......................... 320 310 300 286 272
Tata............................ 321 310 301 287 272
Toyota.......................... 342 333 323 308 294
Volkswagen...................... 341 331 322 307 293
----------------------------------------------------------------------------------------------------------------

These estimates were aggregated based on projected production
volumes into the fleet-wide averages for cars and trucks (Table
III.B.1-3).\171\
---------------------------------------------------------------------------

\171\ Due to rounding during calculations, the estimated fleet-
wide CO2-equivalent levels may vary by plus or minus 1
gram.

Table III.B.1-3--Estimated Fleet-Wide CO2-Equivalent Levels
Corresponding to the Standards
------------------------------------------------------------------------
Cars Trucks
Model year -----------------------------------
CO2 (g/mi) CO2 (g/mi)
------------------------------------------------------------------------
2012................................ 263 346
2013................................ 256 337
2014................................ 247 326
2015................................ 236 312
2016 and later...................... 225 298
------------------------------------------------------------------------

As shown in Table III.B.1-3, fleet-wide CO2-equivalent
emission levels for cars under the approach are projected to decrease
from 263 to 225 grams per mile between MY 2012 and MY 2016. Similarly,
fleet-wide CO2-equivalent

[[Page 25407]]

emission levels for trucks are projected to decrease from 346 to 398
grams per mile. These numbers do not include the effects of other
flexibilities and credits in the program. The estimated achieved values
can be found in Chapter 5 of the Regulatory Impact Analysis (RIA).
EPA has also estimated the average fleet-wide levels for the
combined car and truck fleets. These levels are provided in Table
III.B.1-4. As shown, the overall fleet average CO2 level is
expected to be 250 g/mile in 2016.

Table III.B.1-4--Estimated Fleet-Wide Combined CO2-Equivalent Levels
Corresponding to the Standards
------------------------------------------------------------------------
Combined car
and truck
Model year ---------------
CO2 (g/mi)
------------------------------------------------------------------------
2012.................................................... 295
2013.................................................... 286
2014.................................................... 276
2015.................................................... 263
2016.................................................... 250
------------------------------------------------------------------------

As noted above, EPA is finalizing standards that will result in
increasingly stringent levels of CO2 control from MY 2012
though MY 2016--applying the CO2 footprint curves applicable
in each model year to the vehicles expected to be sold in each model
year produces fleet-wide annual reductions in CO2 emissions.
Comments from the Center for Biological Diversity (CBD) challenged EPA
to increase the stringency of the standards for all of the years of the
program, and even argued that 2016 standards should be feasible in
2012. Other commenters noted the non-linear increase in the standards
from 2011 (CAFE) to the 2012 GHG standards. As explained in greater
detail in Section III.D below and the relevant support documents, EPA
believes that the level of improvement achieves important
CO2 emissions reductions through the application of feasible
control technology at reasonable cost, considering the needed lead time
for this program. EPA further believes that the averaging, banking and
trading provisions, as well as other credit-generating mechanisms,
allow manufacturers further flexibilities which reduce the cost of the
CO2 standards and help to provide adequate lead time. EPA
believes this approach is justified under section 202(a) of the Clean
Air Act.
EPA has analyzed the feasibility under the CAA of achieving the
CO2 standards, based on projections of what actions
manufacturers are expected to take to reduce emissions. The results of
the analysis are discussed in detail in Section III.D below and in the
RIA. EPA also presents the estimated costs and benefits of the car and
truck CO2 standards in Section III.H. In developing the
final rule, EPA has evaluated the kinds of technologies that could be
utilized by the automobile industry, as well as the associated costs
for the industry and fuel savings for the consumer, the magnitude of
the GHG reductions that may be achieved, and other factors relevant
under the CAA.
With respect to the lead time and cost of incorporating technology
improvements that reduce GHG emissions, EPA and NHTSA place important
weight on the fact that during MYs 2012-2016 manufacturers are expected
to redesign and upgrade their light-duty vehicle products (and in some
cases introduce entirely new vehicles not on the market today). Over
these five model years there will be an opportunity for manufacturers
to evaluate almost every one of their vehicle model platforms and add
technology in a cost-effective way to control GHG emissions and improve
fuel economy. This includes redesign of the air conditioner systems in
ways that will further reduce GHG emissions. The time-frame and levels
for the standards, as well as the ability to average, bank and trade
credits and carry a deficit forward for a limited time, are expected to
provide manufacturers the time needed to incorporate technology that
will achieve GHG reductions, and to do this as part of the normal
vehicle redesign process. This is an important aspect of the final
rule, as it will avoid the much higher costs that will occur if
manufacturers needed to add or change technology at times other than
these scheduled redesigns. This time period will also provide
manufacturers the opportunity to plan for compliance using a multi-year
time frame, again in accord with their normal business practice.
Further details on lead time, redesigns and feasibility can be found in
Section III-D.
Consistent with the requirement of CAA section 202(a)(1) that
standards be applicable to vehicles ``for their useful life,'' EPA is
finalizing CO2 vehicle standards that will apply for the
useful life of the vehicle. Under section 202(i) of the Act, which
authorized the Tier 2 standards, EPA established a useful life period
of 10 years or 120,000 miles, whichever first occurs, for all Tier 2
light-duty vehicles and light-duty trucks.\172\ Tier 2 refers to EPA's
standards for criteria pollutants such as NOX, HC, and CO.
EPA is finalizing new CO2 standards for the same group of
vehicles, and therefore the Tier 2 useful life will apply for
CO2 standards as well. The in-use emission standard will be
10% higher than the model-level certification emission test results, to
address issues of production variability and test-to-test variability.
The in-use standard is discussed in Section III.E.
---------------------------------------------------------------------------

\172\ See 65 FR 6698 (February 10, 2000).
---------------------------------------------------------------------------

EPA is requiring manufacturers to measure CO2 for
certification and compliance purposes using the same test procedures
currently used by EPA for measuring fuel economy. These procedures are
the Federal Test Procedure (FTP or ``city'' test) and the Highway Fuel
Economy Test (HFET or ``highway'' test).\173\ This corresponds with the
data used to develop the footprint-based CO2 standards,
since the data on control technology efficiency was also developed in
reference to these test procedures. Although EPA recently updated the
test procedures used for fuel economy labeling, to better reflect the
actual in-use fuel economy achieved by vehicles, EPA is not using these
test procedures for the CO2 standards in this final rule,
given the lack of data on control technology effectiveness under these
procedures.\174\ There were a number of commenters that advocated for a
change in either the test procedures or the fuel economy calculation
weighting factors. The U.S. Coalition for Advanced Diesel Cars urged a
changing of the city/highway weighting factors from their current
values of 45/55 to 43/57 to be more consistent with the EPA (5-cycle)
fuel economy labeling rule. EPA has decided that such a change would
not be appropriate, nor consistent with the technical analyses
supporting the 5-cycle fuel economy label rulemaking. The city/highway
weighting of 43/57 was found to be appropriate when the city fuel
economy is based on a combination of Bags 2 and 3 of the FTP and the
city portion of the US06 test cycle, and when the highway fuel economy
is based on a combination of the HFET and the highway portion of the
US06 cycle. When city and highway fuel economy are based on the FTP and
HFET cycles, respectively, the appropriate city/highway weighting is
not 43/57, but very close to 55/45. Therefore, the weighting of the
city and

[[Page 25408]]

highway fuel economy values contained in this rule is appropriate for
and consistent with the use of the FTP and HFET cycles to measure city
and highway fuel economy.
---------------------------------------------------------------------------

\173\ EPA established the FTP for emissions measurement in the
early 1970s. In 1976, in response to the Energy Policy and
Conservation Act (EPCA) statute, EPA extended the use of the FTP to
fuel economy measurement and added the HFET. The provisions in the
1976 regulation, effective with the 1977 model year, established
procedures to calculate fuel economy values both for labeling and
for CAFE purposes.
\174\ See 71 FR 77872, December 27, 2006.
---------------------------------------------------------------------------

The American Council for an Energy-Efficient Economy (ACEEE),
Cummins, and Sierra Club all suggested using more real-world test
procedures. It is not feasible at this time to base the final
CO2 standards on EPA's five-cycle fuel economy formulae.
Consistent with its name, these formulae require vehicle testing over
five test cycles, the two cycles associated with the proposed
CO2 standards, plus the cold temperature FTP, the US06 high
speed, high acceleration cycle and the SC03 air conditioning test. EPA
considered employing the five-cycle calculation of fuel economy and GHG
emissions for this rule, but there were a number of reasons why this
was not practical. As discussed extensively in the Joint TSD, setting
the appropriate levels of CO2 standards requires extensive
knowledge of the CO2 emission control effectiveness over the
certification test cycles. Such knowledge has been gathered over the
FTP and HFET cycles for decades, but is severely lacking for the other
three test cycles. EPA simply lacks the technical basis to project the
effectiveness of the available technologies over these three test
cycles and therefore, could not adequately support a rule which set
CO2 standards based on the five-cycle formulae. The benefits
of today's rule do presume a strong connection between CO2
emissions measured over the FTP and HFET cycles and onroad operation.
Since CO2 emissions determined by the five-cycle formulae
are believed to correlate reasonably with onroad emissions, this
implies a strong connection between emissions over the FTP and HFET
cycles and the five cycle formulae. However, while we believe that this
correlation is reasonable on average for the vehicle fleet, it may not
be reasonable on a per vehicle basis, nor for any single manufacturer's
vehicles. Thus, we believe that it is reasonable to project a direct
relationship between the percentage change in CO2 emissions
over the two certification cycles and onroad emissions (a surrogate of
which is the five-cycle formulae), but not reasonable to base the
certification of specific vehicles on that untested relationship.
Furthermore, EPA is allowing for off-cycle credits to encourage
technologies that may not be not properly captured on the 2-cycle city/
highway test procedure (although these credits could apply toward
compliance with EPA's standards, not toward compliance with the CAFE
standards). For future analysis, EPA will consider examining new drive
cycles and test procedures for fuel economy.\175\
---------------------------------------------------------------------------

\175\ There were also a number of comments on air conditioner
test procedures; these will be discussed in Section III.C and the
RIA.
---------------------------------------------------------------------------

EPA is finalizing standards that include hydrocarbons (HC) and
carbon monoxide (CO) in its CO2 emissions calculations on a
CO2-equivalent basis. It is well accepted that HC and CO are
typically oxidized to CO2 in the atmosphere in a relatively
short period of time and so are effectively part of the CO2
emitted by a vehicle. In terms of standard stringency, accounting for
the carbon content of tailpipe HC and CO emissions and expressing it as
CO2-equivalent emissions will add less than one percent to
the overall CO2-equivalent emissions level. This will also
ensure consistency with CAFE calculations since HC and CO are included
in the ``carbon balance'' methodology that EPA uses to determine fuel
usage as part of calculating vehicle fuel economy levels.
2. What are the CO2 attribute-based standards?
EPA is finalizing the same vehicle category definitions that are
used in the CAFE program for the 2011 model year standards.\176\ This
approach allows EPA's CO2 standards and the CAFE standards
to be harmonized across all vehicles. In other words, vehicles will be
subject to either car standards or truck standards under both programs,
and not car standards under one program and trucks standards under the
other. The CAFE vehicle category definitions differ slightly from the
EPA definitions for cars and light trucks used for the Tier 2 program
and other EPA vehicle programs. However, EPA is not changing the
vehicle category definitions for any other light-duty mobile source
programs, except the GHG standards.
---------------------------------------------------------------------------

\176\ See 49 CFR 523.
---------------------------------------------------------------------------

EPA is finalizing separate car and truck standards, that is,
vehicles defined as cars have one set of footprint-based curves for MY
2012-2016 and vehicles defined as trucks have a different set for MY
2012-2016. In general, for a given footprint the CO2 g/mi
target for trucks is less stringent then for a car with the same
footprint.
Some commenters requested a single or converging curve for both
cars and trucks.\177\ EPA is not finalizing a single fleet standard
where all cars and trucks are measured against the same footprint curve
for several reasons. First, some vehicles classified as trucks (such as
pick-up trucks) have certain attributes not common on cars which
attributes contribute to higher CO2 emissions--notably high
load carrying capability and/or high towing capability.\178\ Due to
these differences, it is reasonable to separate the light-duty vehicle
fleet into two groups. Second, EPA wishes to harmonize key program
design elements of the GHG standards with NHTSA's CAFE program where it
is reasonable to do so. NHTSA is required by statute to set separate
standards for passenger cars and for non-passenger cars. As discussed
in Section IV, EPCA does not preclude NHTSA from issuing converging
standards if its analysis indicates that these are the appropriate
standards under the statute applicable separately to each fleet.
---------------------------------------------------------------------------

\177\ CBD, ICCT and NESCAUM supported a single curve and the
students at UC Santa Barbara commented on converging curves.
\178\ There is a distinction between body-on-frame trucks and
unibody cars and trucks that make them technically different in a
number of ways. Also, 2WD vehicles tend to have lower CO2
emissions than their 4WD counterparts (all other things being
equal). More discussion of this can be found in the TSD and RIA.
---------------------------------------------------------------------------

Finally, most of the advantages of a single standard for all light
duty vehicles are also present in the two-fleet standards finalized
here. Because EPA is allowing unlimited credit transfer between a
manufacturer's car and truck fleets, the two fleets can essentially be
viewed as a single fleet when manufacturers consider compliance
strategies. Manufacturers can thus choose on which vehicles within
their fleet to focus GHG reducing technology and then use credit
transfers as needed to demonstrate compliance, just as they will if
there was a single fleet standard. The one benefit of a single light-
duty fleet not captured by a two-fleet approach is that a single fleet
prevents potential ``gaming'' of the car and truck definitions to try
and design vehicles which are more similar to passenger cars but which
may meet the regulatory definition of trucks. Although this is of
concern to EPA, we do not believe at this time that concern is
sufficient to outweigh the other reasons for finalizing separate car
and truck fleet standards. However, it is possible that in the future,
recent trends may continue such that cars may become more truck-like
and trucks may become more car-like. Therefore, EPA will reconsider
whether it is appropriate to use converging curves if justified by
future analysis.
For model years 2012 and later, EPA is finalizing a series of
CO2 standards that are described mathematically by a family
of piecewise linear functions

[[Page 25409]]

(with respect to vehicle footprint).\179\ The form of the function is
as follows:
---------------------------------------------------------------------------

\179\ See final regulations at 40 CFR 86.1818-12.

CO2 = a, if x <= l
CO2 = cx + d, if l < x <= h
CO2 = b, if x > h

Where:

CO2 = the CO2 target value for a given
footprint (in g/mi)
a = the minimum CO2 target value (in g/mi)
b = the maximum CO2 target value (in g/mi)
c = the slope of the linear function (in g/mi per sq ft)
d = is the zero-offset for the line (in g/mi CO2)
x = footprint of the vehicle model (in square feet, rounded to the
nearest tenth)
l & h are the lower and higher footprint limits, constraints, or the
boundary (``kinks'') between the flat regions and the intermediate
sloped line

EPA's parameter values that define the family of functions for the
CO2 fleetwide average car and truck standards are as
follows:

Table III.B.2-1--Parameter Values for Cars
[For CO2 gram per mile targets]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Lower Upper
Model year a b c d constraint constraint
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012.................................................... 244 315 4.72 50.5 41 56
2013.................................................... 237 307 4.72 43.3 41 56
2014.................................................... 228 299 4.72 34.8 41 56
2015.................................................... 217 288 4.72 23.4 41 56
2016 and later.......................................... 206 277 4.72 12.7 41 56
--------------------------------------------------------------------------------------------------------------------------------------------------------

Table III.B.2-2--Parameter Values for Trucks
[For CO2 gram per mile targets]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Lower Upper
Model year a b c d constraint constraint
--------------------------------------------------------------------------------------------------------------------------------------------------------
2012.................................................... 294 395 4.04 128.6 41 66
2013.................................................... 284 385 4.04 118.7 41 66
2014.................................................... 275 376 4.04 109.4 41 66
2015.................................................... 261 362 4.04 95.1 41 66
2016 and later.......................................... 247 348 4.04 81.1 41 66
--------------------------------------------------------------------------------------------------------------------------------------------------------

The equations can be shown graphically for each vehicle category,
as shown in Figures III.B.2-1 and III.B.2-2. These standards (or
functions) decrease from 2012-2016 with a vertical shift.
The EPA received a number of comments on both the attribute and the
shape of the curve. For reasons described in Section IIC and Chapter 2
of the TSD, the EPA feels that footprint is the most appropriate choice
of attribute for this rule. More background discussion on other
alternative attributes and curves EPA explored can be found in the EPA
RIA. EPA recognizes that the CAA does not mandate that EPA use an
attribute based standard, as compared to NHTSA's obligations under
EPCA. The EPA believes that a footprint-based program will harmonize
EPA's program and the CAFE program as a single national program,
resulting in reduced compliance complexity for manufacturers. EPA's
reasons for using an attribute based standard are discussed in more
detail in the Joint TSD. Also described in these other sections are the
reasons why EPA is finalizing the slopes and the constraints as shown
above. For future analysis, EPA will consider other options and
suggestions made by commenters.
EPA also received public comments from three manufacturers, General
Motors, Ford Motor Company, and Chrysler, suggesting that the GHG
program should harmonize with an EPCA provision that allows a
manufacturer to exclude emergency vehicles from its CAFE fleet by
providing written notice to NHTSA.\180\ These manufacturers believe
this provision is necessary because law enforcement vehicles (e.g.,
police cars) must be designed with special performance and features
necessary for police work--but which tend to raise GHG emissions and
reduce fuel economy relative to the base vehicle. These commenters
provided several examples of features unique to these special purpose
vehicles that negatively impact GHG emissions, such as heavy-duty
suspensions, unique engine and transmission calibrations, and heavy-
duty components (e.g., batteries, stabilizer bars, engine cooling).
These manufacturers believe consistency in addressing these vehicles
between the EPA and NHTSA programs is critical, as a manufacturer may
be challenged to continue providing the performance needs of the
Federal, State, and local government purchasers of emergency vehicles.
---------------------------------------------------------------------------

\180\ 49 U.S.C. 32902(e).
---------------------------------------------------------------------------

EPA is not finalizing such an emergency vehicle provision in this
rule, since we believe that it is feasible for manufacturers to apply
the same types of technologies to the base emergency vehicle as they
would to other vehicles in their fleet. However, EPA also recognizes
that, because of the unique ``performance upgrading'' needed to convert
a base vehicle into one that meets the performance demands of the law
enforcement community--which tend to reduce GHGs relative to the base
vehicles--there could be situations where a manufacturer is more
challenged in meeting the GHG standards than the CAFE standards, simply
due to inclusion of these higher-emitting vehicles in the GHG program
fleet. While EPA is not finalizing such an exclusion for emergency
vehicles today, we do believe it is important to assess this issue in
the future. EPA plans to assess the unique characteristics of these
emergency vehicles and whether special provisions for addressing them
are warranted. EPA plans to undertake this evaluation as part of a
follow-up rulemaking in the next 18 months (this rulemaking is
discussed in the context of small

[[Page 25410]]

volume manufacturers in Section III.B.6. below).
BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TR07MY10.014

[[Page 25411]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.015

BILLING CODE 6560-50-C

[[Page 25412]]

3. Overview of How EPA's CO2 Standards Will Be Implemented
for Individual Manufacturers
This section provides a brief overview of how EPA will implement
the CO2 standards. Section III.E explains EPA's approach to
certification and compliance in detail. As proposed, EPA is finalizing
two kinds of standards--fleet average standards determined by a
manufacturer's fleet makeup, and in-use standards that will apply to
the individual vehicles that make up the manufacturer's fleet. Although
this is similar in concept to the current light-duty vehicle Tier 2
program, there are important differences. In explaining EPA's
CO2 standards, it is useful to summarize how the Tier 2
program works.
Under Tier 2, manufacturers select a test vehicle prior to
certification and test the vehicle and/or its emissions hardware to
determine both its emissions performance when new and the emissions
performance expected at the end of its useful life. Based on this
testing, the vehicle is assigned to one of several specified bins of
emissions levels, identified in the Tier 2 rule, and this bin level
becomes the emissions standard for the test group the test vehicle
represents. All of the vehicles in the group must meet the emissions
level for that bin throughout their useful life. The emissions level
assigned to the bin is also used in calculating the manufacturer's
fleet average emissions performance.
Since compliance with the Tier 2 fleet average depends on actual
test group sales volumes and bin levels, it is not possible to
determine compliance at the time the manufacturer applies for and
receives a certificate of conformity for a test group. Instead, at
certification, the manufacturer demonstrates that the vehicles in the
test group are expected to comply throughout their useful life with the
emissions bin assigned to that test group, and makes a good faith
demonstration that its fleet is expected to comply with the Tier 2
average when the model year is over. EPA issues a certificate for the
vehicles covered by the test group based on this demonstration, and
includes a condition in the certificate that if the manufacturer does
not comply with the fleet average then production vehicles from that
test group will be treated as not covered by the certificate to the
extent needed to bring the manufacturer's fleet average into compliance
with Tier 2.
EPA is retaining the Tier 2 approach of requiring manufacturers to
demonstrate in good faith at the time of certification that vehicles in
a test group will meet applicable standards throughout useful life. EPA
is also retaining the practice of conditioning certificates upon
attainment of the fleet average standard. However, there are several
important differences between a Tier 2 type of program and the
CO2 standards program. These differences and resulting
modifications to EPA's certification protocols are summarized below and
are described in detail in Section III.E.
EPA will continue to certify test groups as it does for Tier 2, and
the CO2 emission results for the test vehicle will serve as
the initial or default standard for all of the vehicles in the test
group. However, manufacturers will later collect and submit data for
individual vehicle model types \181\ within each test group, based on
the extensive fuel economy testing that occurs through the course of
the model year. This model type data will be used to assign a distinct
certification level for each model type, thus replacing the initial
test group data as the compliance value for each model. It is these
model type values that will be used to calculate the fleet average
after the end of the model year.\182\ The option to substitute model
type data for the test group data is at the manufacturer's discretion,
except they are required, as they are under the CAFE test protocols, to
submit sufficient vehicle test data to represent no less than 90
percent of their actual model year production. The test group emissions
data will continue to apply for any model type that is not covered by
vehicle test data specific to that model type.
---------------------------------------------------------------------------

\181\ ``Model type'' is defined in 40 CFR 600.002-08 as ``* * *
a unique combination of car line, basic engine, and transmission
class.'' A ``car line'' is essentially a model name, such as
``Camry,'' ``Malibu,'' or ``F150.'' The fleet average is calculated
on the basis of model type emissions.
\182\ The final in-use vehicle standards for each vehicle will
also be based on the testing used to determine the model type
values. As discussed in Section III.E.4, an in-use adjustment factor
will be applied to the vehicle test results to determine the in-use
standard that will apply during the useful life of the vehicle.
---------------------------------------------------------------------------

EPA's CO2 standards also differ from Tier 2 in that the
fleet average calculation for Tier 2 is based on test group bin levels
and test group sales whereas under the CO2 program the
CO2 fleet average could be based on a combination of test
group and model type emissions and model type production. For the new
CO2 standards, the final regulations use production rather
than sales in calculating the fleet average in order to closely conform
with the CAFE program, which is a production-based program.\183\
Production as defined in the regulations is relatively easy for
manufacturers to track, but once the vehicle is delivered to
dealerships the manufacturer becomes once step removed from the sale to
the ultimate customer, and it becomes more difficult to track that
final transaction. There is no environmental impact of using production
instead of actual sales, and many commenters supported maintaining
alignment between EPA's program and the CAFE program where possible.
---------------------------------------------------------------------------

\183\ ``Production'' is defined as ``vehicles produced and
delivered for sale'' and is not a measure of the number of vehicles
actually sold.
---------------------------------------------------------------------------

4. Averaging, Banking, and Trading Provisions for CO2
Standards
As explained above, EPA is finalizing a fleet average
CO2 program for passenger cars and light trucks. EPA has
previously implemented similar averaging programs for a range of motor
vehicle types and pollutants, from the Tier 2 fleet average for
NOX to motorcycle hydrocarbon (HC) plus oxides of nitrogen
(NOX) emissions to NOX and particulate matter
(PM) emissions from heavy-duty engines.\184\ The program will operate
much like EPA's existing averaging programs in that manufacturers will
calculate production-weighted fleet average emissions at the end of the
model year and compare their fleet average with a fleet average
emission standard to determine compliance. As in other EPA averaging
programs, the Agency is also finalizing a comprehensive program for
averaging, banking, and trading of credits which together will help
manufacturers in planning and implementing the orderly phase-in of
emissions control technology in their production, consistent with their
typical redesign schedules.\185\
---------------------------------------------------------------------------

\184\ For example, see the Tier 2 light-duty vehicle emission
standards program (65 FR 6698, February 10, 2000), the 2010 and
later model year motorcycle emissions program (69 FR 2398, January
15, 2004), and the 2007 and later model year heavy-duty engine and
vehicle standards program (66 FR 5001, January 18, 2001).
\185\ See final regulations at 40 CFR 86.1865-12.
---------------------------------------------------------------------------

Averaging, Banking, and Trading (ABT) of emissions credits has been
an important part of many mobile source programs under CAA Title II,
both for fuels programs as well as for engine and vehicle programs. ABT
is important because it can help to address many issues of
technological feasibility and lead-time, as well as considerations of
cost. ABT is an integral part of the standard setting itself, and is
not just an add-on to help reduce costs. In many cases, ABT resolves
issues of lead-time

[[Page 25413]]

or technical feasibility, allowing EPA to set a standard that is either
numerically more stringent or goes into effect earlier than could have
been justified otherwise. This provides important environmental
benefits and at the same time it increases flexibility and reduces
costs for the regulated industry. A wide range of commenters expressed
general support for the ABT provisions. Some commenters noted issues
regarding specific provisions of the ABT program, which will be
discussed in the appropriate context below. Several commenters
requested that EPA publicly release manufacturer-specific ABT data to
improve the transparency of credit transactions. These comments are
addressed in Section III.E.
This section discusses generation of credits by achieving a fleet
average CO2 level that is lower than the manufacturer's
CO2 fleet average standard. The final rule includes a
variety of additional ways credits may be generated by manufacturers.
Section III.C describes these additional opportunities to generate
credits in detail. Manufacturers may earn credits through A/C system
improvements beyond a specified baseline. Credits can also be generated
by producing alternative fuel vehicles, by producing advanced
technology vehicles including electric vehicles, plug-in hybrids, and
fuel cell vehicles, and by using technologies that improve off-cycle
emissions. In addition, early credits can be generated prior to the
program's MY 2012 start date. The credits will be used to determine a
manufacturer's compliance at the end of the model year. These credit
generating opportunities are described below in Section III.C.
As explained earlier, manufacturers will determine the fleet
average standard that applies to their car fleet and the standard for
their truck fleet from the applicable attribute-based curve. A
manufacturer's credit or debit balance will be determined by comparing
their fleet average with the manufacturer's CO2 standard for
that model year. The standard will be calculated from footprint values
on the attribute curve and actual production levels of vehicles at each
footprint. A manufacturer will generate credits if its car or truck
fleet achieves a fleet average CO2 level lower than its
standard and will generate debits if its fleet average CO2
level is above that standard. At the end of the model year, each
manufacturer will calculate a production-weighted fleet average for
each averaging set (cars and trucks). A manufacturer's car or truck
fleet that achieves a fleet average CO2 level lower than its
standard will generate credits, and if its fleet average CO2
level is above that standard its fleet will generate debits.
The regulations will account for the difference in expected
lifetime vehicle miles traveled (VMT) between cars and trucks in order
to preserve CO2 reductions when credits are transferred
between cars and trucks. As directed by EISA, NHTSA accomplishes this
in the CAFE program by using an adjustment factor that is applied to
credits when they are transferred between car and truck compliance
categories. The CAFE adjustment factor accounts for two different
influences that can cause the transfer of car and truck credits
(expressed in tenths of a mpg), if left unadjusted, to potentially
negate fuel reductions. First, mpg is not linear with fuel consumption,
i.e., a 1 mpg improvement above a standard will imply a different
amount of actual fuel consumed depending on the level of the standard.
Second, NHTSA's conversion corrects for the fact that the typical
lifetime miles for cars is less than that for trucks, meaning that
credits earned for cars and trucks are not necessarily equal. NHTSA's
adjustment factor essentially converts credits into vehicle lifetime
gallons to ensure preservation of fuel savings and the transfer credits
on an equal basis, and then converts back to the statutorily-required
credit units of tenths of a mile per gallon. To convert to gallons
NHTSA's conversion must take into account the expected lifetime mileage
for cars and trucks. Because EPA's standards are expressed on a
CO2 gram per mile basis, which is linear with fuel
consumption, EPA's credit calculations do not need to account for the
first issue noted above. However, EPA is accounting for the second
issue by expressing credits when they are generated in total lifetime
Megagrams (metric tons), rather than through the use of conversion
factors that would apply at certain times. In this way credits may be
freely exchanged between car and truck compliance categories without
the need for adjustment. Additional detail regarding this approach,
including a discussion of the vehicle lifetime mileage estimates for
cars and trucks can be found in Section III.E.5. A discussion of the
derivation of the estimated vehicle lifetime miles traveled can be
found in Chapter 4 of the Joint Technical Support Document.
A manufacturer that generates credits in a given year and vehicle
category may use those credits in essentially four ways, although with
some limitations. These provisions are very similar to those of other
EPA averaging, banking, and trading programs. These provisions have the
potential to reduce costs and compliance burden, and support the
feasibility of the standards in terms of lead time and orderly redesign
by a manufacturer, thus promoting and not reducing the environmental
benefits of the program.
First, EPA proposed that the manufacturer must use any credits
earned to offset any deficit that had accrued in the current year or in
a prior model year that had been carried over to the current model
year. NRDC commented that such a provision is necessary to prevent
credit ``shell games'' from delaying the adoption of new technologies.
EPA's Tier 2 program includes such a restriction, and EPA is applying
an identical restriction to the GHG program. Simply stated, a
manufacturer may not bank (or carry forward) credits if that
manufacturer is also carrying a deficit. In such a case, the
manufacturer is obligated to use any current model year credits to
offset that deficit. Using current model year credits to offset a prior
model year deficit is referred to in the CAFE program as credit carry-
back. EPA's deficit carry-forward, or credit carry-back provisions are
described further, below.
Second, after satisfying any needs to offset pre-existing deficits,
remaining credits may be banked, or saved for use in future years.
Credits generated in this program will be available to the manufacturer
for use in any of the five model years after the model year in which
they were generated, consistent with the CAFE program under EISA. This
is also referred to as a credit carry-forward provision.
EPA received a number of comments regarding the credit carry-back
and carry-forward provisions. Many supported the proposed consistency
of these provisions with EISA and the flexibility provided by these
provisions, and several offered qualified or tentative support. For
example, NRDC encouraged EPA to consider further restrictions in the
2017 and later model years. Public Citizen expressed concern regarding
the complexity of the program and how these provisions might obscure a
straightforward determination of compliance in any given model year. At
least two automobile manufacturers suggested modeling the program after
California, which allows credits to be carried forward for three
additional years following a discounting schedule.
For other new emission control programs, EPA has sometimes
initially restricted credit life to allow time for the Agency to assess
whether the credit program is functioning as intended. When EPA first
offered averaging and

[[Page 25414]]

banking provisions in its light-duty emissions control program (the
National Low Emission Vehicle Program), credit life was restricted to
three years. The same is true of EPA's early averaging and banking
program for heavy-duty engines. As these programs matured and were
subsequently revised, EPA became confident that the programs were
functioning as intended and that the standards were sufficiently
stringent to remove the restrictions on credit life. EPA is therefore
acting consistently with our past practice in finalizing reasonable
restrictions on credit life in this new program. The Agency believes
that a credit life of five years represents an appropriate balance
between promoting orderly redesign and upgrade of the emissions control
technology in the manufacturer's fleet and the policy goal of
preventing large numbers of credits accumulated early in the program
from interfering with the incentive to develop and transition to other
more advanced emissions control technologies. As discussed below in
Section III.C, early credits generated by a manufacturer are also be
subject to the five year credit carry-forward restriction based on the
year in which they are generated. This limits the effect of the early
credits on the long-term emissions reductions anticipated to result
from the new standards.
Third, the new program enables manufacturers to transfer credits
between the two averaging sets, passenger cars and trucks, within a
manufacturer. For example, credits accrued by over-compliance with a
manufacturer's car fleet average standard may be used to offset debits
accrued due to that manufacturer's not meeting the truck fleet average
standard in a given year. EPA believes that such cross-category use of
credits by a manufacturer provides important additional flexibility in
the transition to emissions control technology without affecting
overall emission reductions. Comments regarding the credit transfer
provisions expressed general support, noting that it does not matter to
the environment whether a gram of greenhouse gas is generated from a
car or a truck. Additional comments regarding EPA's streamlined
megagram approach and method of accounting for expected vehicle
lifetime miles traveled are summarized in Section III.E.
Finally, accumulated credits may be traded to another vehicle
manufacturer. As with intra-company credit use, such inter-company
credit trading provides flexibility in the transition to emissions
control technology without affecting overall emission reductions.
Trading credits to another vehicle manufacturer could be a
straightforward process between the two manufacturers, but could also
involve third parties that could serve as credit brokers. Brokers may
not own the credits at any time. These sorts of exchanges are typically
allowed under EPA's current emission credit programs, e.g., the Tier 2
light-duty vehicle NOX fleet average standard and the heavy-
duty engine NOX fleet average standards, although
manufacturers have seldom made such exchanges. Comments generally
reflected support for the credit trading flexibility, although some
questioned the extent to which trading might actually occur. As noted
above, comments regarding program transparency are addressed in Section
III.E.
If a manufacturer has accrued a deficit at the end of a model
year--that is, its fleet average level failed to meet the required
fleet average standard--the manufacturer may carry that deficit forward
(also referred to credit carry-back) for a total of three model years
after the model year in which that deficit was generated. EPA continues
to believe that three years is an appropriate amount of time that gives
the manufacturers adequate time to respond to a deficit situation but
does not create a lengthy period of prolonged non-compliance with the
fleet average standards.\186\ As noted above, such a deficit carry-
forward may only occur after the manufacturer has applied any banked
credits or credits from another averaging set. If a deficit still
remains after the manufacturer has applied all available credits, and
the manufacturer did not obtain credits elsewhere, the deficit may be
carried forward for up to three model years. No deficit may be carried
into the fourth model year after the model year in which the deficit
occurred. Any deficit from the first model year that remains after the
third model year will constitute a violation of the condition on the
certificate, which will constitute a violation of the Clean Air Act and
will be subject to enforcement action.
---------------------------------------------------------------------------

\186\ EPA emission control programs that incorporate ABT
provisions (e.g., the Tier 2 program and the Mobile Source Air
Toxics program) have provided this three-year deficit carry-forward
provision for this reason. See 65 FR 6745 (February 10, 2000), and
71 FR 8427 (February 26, 2007).
---------------------------------------------------------------------------

The averaging, banking, and trading provisions are generally
consistent with those included in the CAFE program, with a few notable
exceptions. As with EPA's approach, CAFE allows five year carry-forward
of credits and three year carry-back. Under CAFE, transfers of credits
across a manufacturer's car and truck averaging sets are also allowed,
but with limits established by EISA on the use of transferred credits.
The amount of transferred credits that can be used in a year is
limited, and transferred credits may not be used to meet the CAFE
minimum domestic passenger car standard. CAFE allows credit trading,
but again, traded credits cannot be used to meet the minimum domestic
passenger car standard. EPA did not propose, and is not finalizing,
these constraints on the use of transferred credits.
Additional details regarding the averaging, banking, and trading
provisions and how EPA will implement these provisions can be found in
Section III.E.
5. CO2 Temporary Lead-Time Allowance Alternative Standards
EPA proposed adopting a limited and narrowly prescribed option,
called the Temporary Lead-time Allowance Alternative Standards (TLAAS),
to provide additional lead time for a certain subset of manufacturers.
As noted in the proposal, this option was designed to address two
different situations where we project that more lead time is needed,
based on the level of emissions control technology and emissions
control performance currently exhibited by certain vehicles. One
situation involves manufacturers who have traditionally paid CAFE fines
instead of complying with the CAFE fleet average, and as a result at
least part of their vehicle production currently has significantly
higher CO2 and lower fuel economy levels than the industry
average. More lead time is needed in the program's initial years to
upgrade these vehicles to meet the aggressive CO2 emissions
performance levels required by the final rule. The other situation
involves manufacturers who have a limited line of vehicles and are
therefore unable to average emissions performance across a full line of
production. For example, some smaller volume manufacturers produce only
vehicles with emissions above the corresponding CO2
footprint target, and do not have other types of vehicles (that exceed
their compliance targets) in their production mix with which to
average. Often, these manufacturers also pay fines under the CAFE
program rather than meeting the applicable CAFE standard. Because
voluntary non-compliance through payment of civil penalties is
impermissible for the GHG standards under the CAA, both of these types
of manufacturers need additional lead time to upgrade vehicles and meet
the standards. EPA proposed that this subset of manufacturers be
allowed to

[[Page 25415]]

produce up to 100,000 vehicles over model years 2012-2015 that would be
subject to a somewhat less stringent CO2 standard of 1.25
times the standard that would otherwise apply to those vehicles. Only
manufacturers with total U.S. sales of less than 400,000 vehicles per
year in MY 2009 would be eligible for this allowance. Those
manufacturers would have to exhaust designated program flexibilities in
order to be eligible, and credit generating and trading opportunities
for the eligible vehicles would be restricted. See 74 FR 49522-224.
EPA is finalizing the optional TLAAS provisions, with certain
limited modifications, so that these manufacturers can have sufficient
lead time to meet the tougher MY 2016 GHG standards, while preserving
consumer choice of vehicles during this time.\187\ EPA is finalizing
modified provisions to address the unique lead-time issues of smaller
volume manufacturers. One provision involves additional flexibility
under the TLAAS program for manufacturers below 50,000 U.S. vehicle
sales, as discussed further in Section III.B.5.b below. Another
provision defers the CO2 standards for the smallest volume
manufacturers, those below 5,000 U.S. vehicle sales, as discussed in
Section III.B.6.
---------------------------------------------------------------------------

\187\ See final regulations at 40 CFR 86.1818-12(e).
---------------------------------------------------------------------------

Comments from several manufacturers strongly supported the TLAAS
program as critical to provide the lead time needed for manufacturers
to meet the standards. Volkswagen commented that TLAAS is an important
aspect of EPA's proposal and that it responds to the needs of some
smaller manufacturers for additional lead time and flexibility under
the CAA. Daimler Automotive Group commented that TLAAS is a critical
element of the program and falls squarely within EPA's discretion to
provide appropriate lead time to limited-line low-volume manufacturers.
BMW also commented that TLAAS is needed because most of the companies
with limited lines will have to meet a more stringent fleet standard by
2016 than full-line manufacturers because they sell ``feature-dense''
vehicles (as opposed to light-weight large wheel-base vehicles) and no
pick-up trucks. BMW commented that their MY 2016 footprint-based
standard is projected to be 4 percent more stringent than the fleet
average standard of 250 g/mile. The Alliance of Automobile
Manufacturers supported the flexibilities proposed by EPA, including
TLAAS. As discussed in detail below, EPA received extensive comments
from many smaller volume manufacturers that the proposed TLAAS program
was insufficient to address lead time and feasibility issues they will
face under the program.
In contrast, EPA also received comments from the Center for
Biological Diversity opposing the TLAAS program, commenting that an
exception for high performance vehicles is not allowed under EPCA or
the CAA and that it rewards manufacturers that pay penalties under CAFE
and penalizes those that have complied with CAFE. This commenter
suggests that manufacturers could decrease vehicle mass or power output
of engines, purchase credits from another manufacturer, or earn off-
cycle credits. EPA responds to these comments below.
After carefully considering the public comments, EPA continues to
believe that the TLAAS program is essential in providing necessary lead
time and flexibility to eligible manufacturers in the early years of
the standards. First, EPA believes that it is acting well within its
legal authority in adopting the various TLAAS provisions. EPA is
required to provide sufficient lead time for industry as a whole for
standards under section 202(a)(1), which mandates that standards are to
take effect only ``after providing such period as the Administrator
finds necessary to permit the development and application of the
requisite technology, giving appropriate consideration to the cost of
compliance within such period.'' Thus, although section 202(a)(1) does
not explicitly authorize this or any other specific lead time
provision, it affords ample leeway for EPA to craft provisions designed
to provide adequate lead time, and to tailor those provisions as
appropriate. We show below that the types of technology penetrations
required for TLAAS-eligible vehicles in the program's earlier years
raise critical issues as to adequacy of lead time. As discussed in the
EPA feasibility analysis provided in Section III.D.6 and III.D.7
several manufacturers eligible for TLAAS are projected to face a
compliance shortfall in MY 2016 without the TLAAS program, even with
the full application of technologies assumed by the OMEGA Model,
including hybrid use of up to 15 percent. These include BMW, Jaguar
Land Rover, Daimler, Porsche, and Volkswagen In addition, the smaller
volume manufacturers of this group (i.e., Jaguar Land Rover and
Porsche) face the greatest shortfall (see Table III.D.6-4). Even with
TLAAS, these manufacturers will need to take technology steps to comply
with standards above and beyond those of other manufacturers. These
manufacturers have relatively few models with high baseline emissions
and this flexibility allows them additional lead time to adapt to a
longer term strategy of meeting the final standards within their
vehicle redesign cycles.
Second, EPA has carefully evaluated other means of eligible
manufacturers to meet the standards, such as utilizing available credit
opportunities. Indeed, eligibility for the TLAAS, and for temporary
deferral of regulation for very small volume manufacturers, is
conditioned on first exhausting the various programmatic flexibilities
including credit utilization. At the same time, a basic reason certain
manufacturers are faced with special lead time difficulties is their
inability to generate credits which can be then be averaged across
their fleet because of limited product lines. And although purchasing
credits is an option under the program, there are no guarantees that
credits will be available. Historic practice in fact suggests that
manufacturers do not sell credits to competitors. While some of the
smaller manufacturers covered by the TLAAS program may be in a position
to obtain credits, they are not likely to be available for the TLAAS
manufacturers across the board in the volume needed to comply without
the TLAAS provisions. At the same time the TLAAS provisions have been
structured such that any credits that do become available would likely
be used before a manufacturer would turn to the more restricted and
limiting TLAAS provisions.
As discussed in Section III.C., off-cycle credits are available if
manufacturers are able to employ new and innovative technologies not
already in widespread use, which provide real-world emissions
reductions not captured on the current test cycles. Further, these
credits are eligible only for technologies that are newly introduced on
just a few vehicle models, and are not yet in widespread use across the
fleet. The magnitude of these credits are highly uncertain because they
are based on new technologies, and EPA is not aware of any such
technologies that would provide enough credits to bring these
manufacturers into compliance without TLAAS lead time flexibility.
Manufacturers first must develop these technologies and then
demonstrate their emissions reductions capabilities, which will require
lead time. Moreover, the technologies mentioned in the proposal which
are the most likely to be eligible based on present knowledge,
including solar panels and active

[[Page 25416]]

aerodynamics, are likely to provide only small incremental emissions
reductions.
We agree with the comment that reducing vehicle mass or power are
potential methods for reducing emissions that should be employed by
TLAAS-eligible manufacturers to help them meet standards. However,
based on our assessment of the lead time needed for these manufacturers
to comply with the standards, especially given their more limited
product offerings and higher baseline emissions, we believe that
additional time is needed for them to come into compliance. EPA can
permissibly consider the TLAAS and other manufacturers' lead time,
cost, and feasibility issues in developing the primary standards and
has discretion in setting the overall stringency of the standards to
account for these factors. Natural Resources Defense Council v. Thomas,
805 F. 2d 410, 421 (DC Cir. 1986) (even when implementing technology-
forcing provisions of Title II, EPA may base standards on an industry-
wide capability ``taking into account the broad spectrum of
technological capabilities as well as cost and other factors'' across
the industry). EPA is not legally required to set standards that drive
these manufacturers or their products out of the market, nor is EPA
legally required to preserve a certain product line or vehicle
characteristic. Instead EPA has broad discretion under section
202(a)(1) to set standards that reasonably balance lead time needs
across the industry as a whole and vehicle availability. In this
rulemaking, EPA has consistently emphasized the importance of obtaining
very significant reductions in emissions of GHGs from the industry as a
whole, and obtaining those reductions through regulatory approaches
that avoid limiting the ability of manufacturers to provide model
availability and choice for consumers. The primary mechanism to achieve
this is the use of a footprint attribute curve in setting the
increasingly stringent model year standards. The TLAAS provisions are a
temporary and strictly limited modification to these attribute
standards allowing the TLAAS manufacturers lead time to upgrade their
product lines to meet the 2016 GHG standards. EPA has made a reasonable
choice here to preserve the overall stringency of the program, and to
afford increased flexibility in the program's early years to a limited
class of vehicles to assure adequate lead time for all manufacturers to
meet the strictest of the standards by MY 2016.
As described below, EPA also carefully considered the comments of
smaller volume manufacturers and believes additional lead time is
needed. Therefore, EPA is finalizing the TLAAS program, similar to that
proposed, and is also finalizing an additional TLAAS option for
manufacturers with annual U.S. sales under 50,000 vehicles. EPA is also
deferring standards for manufacturers with annual sales of less than
5,000 vehicles. These new TLAAS provisions and the small volume
manufacturer deferment are discussed in detail below and in Section
III.B.6.
a. Base TLAAS Program
As proposed, EPA is establishing the TLAAS program for a specified
subset of manufacturers. This alternative standard is an option only
for manufacturers with total U.S. sales of less than 400,000 vehicles
per year, using 2009 model year final sales numbers to determine
eligibility for these alternative standards. For manufacturers with
annual U.S. sales of 50,000 or more but less than 400,000 vehicles, EPA
is finalizing the TLAAS program largely as proposed. EPA proposed that
under the TLAAS, qualifying manufacturers would be allowed to produce
up to 100,000 vehicles that would be subject to a somewhat less
stringent CO2 standard of 1.25 times the standard that would
otherwise apply to those vehicles. This 100,000 volume is not an annual
limit, but is an absolute limit for the total number of vehicles which
can use the TLAAS program over the model years 2012-2015. Any
additional production would be subject to the same standards as any
other manufacturer. EPA is retaining this limit for manufacturers with
baseline MY 2009 sales of 50,000 but less than 400,000. In addition, as
discussed further below, EPA is finalizing a variety of restrictions on
the use of the TLAAS program, to ensure that only manufacturers who
need more lead time for the kinds of reasons noted above are likely to
use the program.
Volvo and Saab commented that basing eligibility strictly on MY
2009 sales would be problematic for these companies, which are being
spun-off from larger manufacturer in the MY 2009 time frame due to the
upheaval in the auto industry over the past few years. These commenters
offered a variety of suggestions including using MY 2010 as the
eligibility cut-off instead of MY 2009, reassessing eligibility on a
year-by-year basis as corporate relationships change, or allowing
companies separated from a larger parent company by the end of 2010 to
use their MY 2009 branded U.S. sales to qualify for TLAAS. In response
to these concerns, EPA recognizes that these companies currently being
sold by larger manufacturers will share the same characteristics of the
manufacturers for which the TLAAS program was designed. As newly
independent companies, these firms will face the challenges of a
narrower fleet of vehicles across which to average, and may potentially
be in a situation, at least in the first few years, of paying fines
under CAFE. Lead time concerns in the program's initial years are in
fact particularly acute for these manufacturers since they will be
newly independent, and thus would have even less of an opportunity to
modify their vehicles to meet the standards. Therefore, EPA is
finalizing an approach that allows manufacturers with U.S. ``branded
sales'' in MY 2009 under the umbrella of a larger manufacturer that
become independent by the end of calendar year 2010 to use their MY
2009 branded sales to qualify for TLAAS eligibility. In other words, a
manufacturer will be eligible for TLAAS if it produced vehicles for the
U.S. market in MY 2009, its branded sales of U.S. vehicles were less
than 400,000 in MY 2009 but whose vehicles were sold as part of a
larger manufacturer, and it becomes independent by the end of calendar
year 2010, if the new entity has sales below 400,000 vehicles.
Manufacturers with no U.S. sales in MY 2009 are not eligible to
utilize the TLAAS program. EPA does not support the commenter's
suggestion of a year-by-year eligibility determination because it opens
up the TLAAS program to an unknown universe of potential eligible
manufacturers, with the potential for gaming. EPA does not believe the
TLAAS program should be available to new entrants to the U.S. market
since these manufacturers are not transitioning from the CAFE regime
which allows fine paying as a means of compliance to a CAA regime which
does not, and hence do not present the same types of lead time issues.
Manufacturers entering the U.S. market for the first time thus will be
fully subject to the GHG fleet-average standards.
As proposed, manufacturers qualifying for TLAAS will be allowed to
meet slightly less stringent standards for a limited number of
vehicles. An eligible manufacturer could have a total of up to 100,000
units of cars or trucks combined over model years 2012-2015 which would
be subject to a standard 1.25 times the standard that would otherwise
apply to those vehicles under the primary program. In other words, the
footprint curves upon which the individual manufacturer standards for
the TLAAS fleets are based would be

[[Page 25417]]

less stringent by a factor of 1.25 for up to 100,000 of an eligible
manufacturer's vehicles for model years 2012-2015. EPA believes that
100,000 units over four model years achieves an appropriate balance, as
the emissions impact is quite small, but does provide companies with
necessary lead time during MY 2012-2015. For example, for a
manufacturer producing 400,000 vehicles per year, this would be a total
of up to 100,000 vehicles out of a total production of up to 1.6
million vehicles over the four year period, or about 6 percent of total
production.
Finally, for manufacturers of 50,000 but less than 400,000 U.S.
vehicles sales during 2009, the program expires at the end of MY 2015
as proposed. EPA continues to believe the program reasonably addresses
a real world lead time constraint for these manufacturers, and does so
in a way that balances the need for more lead time with the need to
minimize any resulting loss in potential emissions reductions. In MY
2016, the TLAAS option thus ends for all but the smallest manufacturers
opting for TLAAS, and manufacturers must comply with the same
CO2 standards as non-TLAAS manufacturers; under the CAFE
program companies would continue to be allowed to pay civil penalties
in lieu of complying with the CAFE standards. However, because
companies must meet both the CAFE standards and the EPA CO2
standards, the National Program will have the practical impact of
providing a level playing field for almost all except the smallest
companies beginning in MY 2016. This option, even with the
modifications being adopted, thereby results in more fuel savings and
CO2 reductions than would be the case under the CAFE program
by itself.
EPA proposed that manufacturers meeting the cut-point of below
400,000 sales for MY 2009 but whose U.S. sales grew above 400,000 in
any subsequent model years would remain eligible for the TLAAS program.
The total sales number applies at the corporate level, so if a
corporation owns several vehicle brands the aggregate sales for the
corporation must be used. These provisions would help prevent gaming of
the provisions through corporate restructuring. Corporate ownership or
control relationships would be based on determinations made under CAFE
for model year 2009 (except in the case of a manufacturer being sold by
a larger manufacturer by the end of calendar year 2010, as discussed
above). In other words, corporations grouped together for purposes of
meeting CAFE standards in MY 2009, must be grouped together for
determining whether or not they are eligible under the 400,000 vehicle
cut point. EPA is finalizing these provisions with the following
modifications. EPA recognizes the dynamic corporate restructuring
occurring in the auto industry and believes it is important to
structure additional provisions to ensure there is no ability to game
the TLAAS provisions and to ensure no unintended loss of feasible
environmental benefits. Therefore, EPA is finalizing a provision that
if two or more TLAAS eligible companies are later merged, with one
company having at least 50% or more ownership of the other, or if the
companies are combined for the purposes of EPA certification and
compliance, the TLAAS allotment is not additive. The merged company
will only be allowed the allotment for what is considered the parent
company under the new corporate structure. Further, if the newly formed
company would have exceeded the 400,000 vehicle cut point based on
combined MY 2009 sales, the new entity is not eligible for TLAAS in the
model year following the merger. EPA believes that such mergers and
acquisitions would give the parent company additional opportunities to
average across its fleet, eliminating one of the primary needs for the
TLAAS program. This provision will not be retroactive and will not
affect the TLAAS program in the year of the merger or for previous
model years. EPA believes these additional provisions are essential to
ensure the integrity of the TLAAS program by ensuring that it does not
become available to large manufacturers through mergers and
acquisitions.
As proposed, the TLAAS vehicles will be separate car and truck
fleets for that model year and subject to the less stringent footprint-
based standards of 1.25 times the primary fleet average that would
otherwise apply. The manufacturer will determine what vehicles are
assigned to these separate averaging sets for each model year. As
proposed, credits from the primary fleet average program can be
transferred and used in the TLAAS program. Credits generated within the
TLAAS program may also be transferred between the TLAAS car and truck
averaging sets (but not to the primary fleet as explained below) for
use through MY 2015 when the TLAAS ends.
EPA is finalizing a number of restrictions on credit trading within
the TLAAS program, as proposed. EPA is concerned that if credit use in
the TLAAS program were unrestricted, some manufacturers would be able
to place relatively clean vehicles in the TLAAS fleet, and generate
credits for the primary program fleet. First, credits generated under
TLAAS may not be transferred or traded to the primary program.
Therefore, any unused credits under TLAAS expire after model year 2015
(or 2016 for manufacturers with annual sales less than 50,000
vehicles). EPA believes that this is necessary to limit the program to
situations where it is needed and to prevent the allowance from being
inappropriately transferred to the long-term primary program where it
is not needed. EPA continues to believe this provision is necessary to
prevent credits from being earned simply by removing some high-emitting
vehicles from the primary fleet. Absent this restriction, manufacturers
would be able to choose to use the TLAAS for these vehicles and also be
able to earn credits under the primary program that could be banked or
traded under the primary program without restriction. Second, EPA is
finalizing two additional restrictions on the use of TLAAS by requiring
that for any of the 2012-2015 model years for which an eligible
manufacturer would like to use the TLAAS, the manufacturer must use two
of the available flexibilities in the GHG program first in order to try
and comply with the primary standard before accessing the TLAAS--i.e.,
TLAAS eligibility is not available to those manufacturers with other
readily-available means of compliance. Specifically, before using the
TLAAS a manufacturer must: (1) Use any banked emission credits from
previous model years; and, (2) use any available credits from the
companies' car or truck fleet for the specific model year (i.e., use
credit transfer from cars to trucks or from trucks to cars). That is,
before using the TLAAS for either the car fleet or the truck fleet, the
company must make use of any available intra-manufacturer credit
transfers first. Finally, EPA is restricting the use of banking and
trading between companies of credits in the primary program in years in
which the TLAAS is being used. No such restriction is in place for
years when the TLAAS is not being used.
EPA received several comments in support of these credit
restrictions for the TLAAS program. On the negative side, one
manufacturer commented that the restrictions were not necessary, saying
that the restrictions are counter to providing manufacturers with
flexibility and that the emissions impacts estimated by EPA due to the
full use of the program are small. However, EPA continues to believe
that the restrictions are appropriate to prevent the potential gaming
described above, and to ensure that the TLAAS

[[Page 25418]]

program is used only by those manufacturers that have exhausted all
other readily available compliance mechanisms and consequently have
legitimate lead time issues.
One manufacturer commented that the program is restrictive due to
the requirement that manufacturers must decide prior to the start of
the model year whether or not and how to use the TLAAS program. EPA did
not intend for manufactures to have to make this determination prior to
the start of the model year. EPA expects that manufacturers will
provide a best estimate of their plans to use the TLAAS program during
certification based on projected model year sales, as part of their pre
model year report projecting their overall plan for compliance (as
required by Sec. 600.514-12 of the regulations). Manufacturers must
determine the program's actual use at the end of the model year during
the process of demonstrating year-end compliance. EPA recognizes that
depending on actual sales for a given model year, a manufacturer's use
of TLAAS may change from the projections used in the pre-model year
report.
b. Additional TLAAS Flexibility for Manufacturers With MY 2009 Sales of
Less Than 50,000 Vehicles
EPA received extensive comments that the TLAAS program would not
provide sufficient lead time and flexibility for companies with sales
of significantly less than 400,000 vehicles. Jaguar Land Rover, which
separated from Ford in 2008, commented that it sells products only in
the middle and large vehicle segments and that its total product range
remains significantly more limited in terms of segments in comparison
with its main competitors which typically have approximately 75% of
their passenger car fleet in the small and middle segments. Jaguar Land
Rover also commented that it has already committed $1.3 billion of
investment to reducing CO2 from its vehicle fleet and that
this investment is already delivering a range of technologies to
improve the fuel economy and CO2 performance of its existing
vehicles. Jaguar Land Rover submitted confidential business information
regarding their future product plans and emissions performance
capabilities of their vehicles which documents their assertions.
Porsche commented that their passenger car footprint-based standard
is the most stringent of any manufacturer and this, combined with their
high baseline emissions level, means that it would need to reduce
emissions by about 10 percent per year over the 2012-2016 time-frame.
Porsche commented that such reductions were not feasible. They
commented that their competitors will be able to continue to offer
their full line of products because the competitors have a wider range
of products with which to average. Porsche further commented that their
product development cycles are longer than larger competitors. Porsche
recommended for small limited line niche manufacturers that EPA require
an annual 5 percent reduction in emissions from baseline up to a total
reduction of 25 percent, or to modify the TLAAS program to require such
reductions. Porsche noted that this percent reduction would be in line
with the average emissions reductions required for larger
manufacturers.
EPA also received comments from several very small volume
manufacturers that, even with the TLAAS program, the proposed standards
are not feasible for them, certainly not in the MY 2012-2016 MY time
frame. These manufacturers included Aston Martin, McLaren, Lotus, and
Ferrari. Their comments consistently focused on the need for separate,
less stringent standards for small volume manufacturers. The
manufacturers commented that they are willing to make progress in
reducing emissions, but that separate, less-stringent small volume
manufacturer standards are needed for them to remain in the U.S.
market. The commenters note that their product line consists entirely
of high end sports cars. Most of these manufacturers have only a few
vehicle models, have annual sales on the order of a few hundred to a
few thousand vehicles, and several have average baseline CO2
emissions in excess of 500 g/mile--nearly twice the industry average.
McLaren commented that its vehicle model to be introduced in MY 2011
will have class leading CO2 performance but that it would
not be able to offer the vehicle in the U.S. market because it does not
have other vehicle models with which to average. Similarly, Aston
Martin commented that it is of utmost importance that it is not
required to reduce emissions significantly more than equivalent
vehicles from larger manufacturers, which would render them
uncompetitive due purely to the size of its business. Manufacturers
also noted that they launch new products less frequently than larger
manufacturers (e.g., Ferrari noted that their production period for
models is 7-8 years), and that suppliers serve large manufacturers
first because they can buy in larger volumes. Some manufacturers also
noted that they would be willing to purchase credits at a reasonable
price, but they believed that credit availability from other
manufacturers was highly unlikely due to the competitive nature of the
auto industry. Several of these manufacturers provided confidential
business information indicating their preliminary plans for reducing
GHG emissions across their product lines through MY 2016 and beyond.
The Association of International Automobile Manufacturers (AIAM)
also commented that, because of their essential features, vehicles
produced by small volume manufacturers would not be able to meet the
proposed greenhouse gas standards. AIAM commented that ``while it is
possible that these small volume manufacturers (SVMs) might be able to
comply with greenhouse gas standards by purchasing credits from other
manufacturers, this is far too speculative a solution. The market for
credits is unpredictable at this point. Other than exiting the U.S.
market, therefore, the only other possible solution for an independent
SVM would be to sell an equity interest in the company to a larger,
full-line manufacturer, so that the emissions of the luxury vehicles
could be averaged in with the much larger volume of other vehicles
produced by the major manufacturer. This cannot possibly be the outcome
EPA intends, especially when measured against the minimal, if any,
environmental benefit that would result.'' AIAM commented further that
``there is ample legal authority for EPA to provide SVMs a more
generous lead-time allowance or an alternative standard. Indeed, EPA
recognizes such authority in the proposal for a small entity exemption
(for those companies defined under the Small Business Administration's
regulations), see 74 FR at 49574, and in the TLAAS. These provisions
are consistent with previous EPA rulemaking under the Clean Air Act
which offer relief to SVMs.'' AIAM recommended deferring standards for
SVMs to a future rulemaking, providing EPA with adequate time to assess
relevant product plans and technology feasibility information from
SVMs, conduct the necessary reviews and modeling that may be needed,
and consult with the stakeholders.
These commenters noted that standards for the smallest
manufacturers were deferred in the California program until MY 2016 and
that California's program would have established standards for small
volume manufacturers in MY 2016 at a level that would be
technologically feasible.

[[Page 25419]]

The commenters also suggested that California's approach is similar to
the approach being taken by EPA for small business entities. Further,
these commenters noted that in Tier 2 and other light-duty vehicle
programs, EPA has allowed small volume manufacturers (SVMs) until the
end of the phase-in period to comply with standards. The commenters
recommended that EPA should defer standards for SVMs, and conduct a
future rulemaking to establish appropriate standards for SVMs starting
in model year 2016. Alternatively, some manufacturers recommended
establishing much less stringent standards for SVMs as part of the
current rulemaking.
In summary, the manufacturers commented that their range of
products was insufficient to allow them to meet the standards in the
time provided, even with the proposed TLAAS program. Many of these
manufacturers have baseline emissions significantly higher than their
larger-volume competitors, and thus the CO2 reductions
required from baseline under the program are larger for many of these
companies than for other companies. Although they are investing
substantial resources to reduce CO2 emissions, they believe
that they will not be able to achieve the standards under the proposed
approach.
EPA also received comments urging us not to expand the TLAAS
program. The commenters are concerned about the loss of benefits that
would occur with any expansion.
EPA has considered the comments carefully and concludes that
additional flexibility is needed for these companies. After assessing
the issues raised by commenters, EPA believes there are two groups of
manufacturers that need additional lead time. The first group includes
manufacturers with annual U.S. sales of less than 5,000 vehicles per
year. Standards for these small volume manufacturers are being deferred
until a future rulemaking in the 2012 timeframe, as discussed in
Section III.B.6, below. This will allow EPA to determine the
appropriate level of standards for these manufacturers, as well as the
small business entities, at a later time. The second group includes
manufacturers with MY 2009 U.S. sales of less than 50,000 vehicles but
above the 5,000 vehicle threshold being established for small volume
manufacturers. EPA has selected a cut point of 50,000 vehicles in order
to limit the additional flexibility to only the smaller manufacturers
with much more limited product lines over which to average. EPA has
tailored these provisions as narrowly as possible to provide additional
lead time only as needed by these smaller manufacturers. We estimate
that the TLAAS program, including the changes below will result in a
total decrease in overall emissions reductions of about one percent of
the total projected GHG program emission benefits. These estimates are
provided in RIA Chapter 5 Appendix A.
For some of the companies, the reduction from baseline
CO2 emissions required to meet the standards is clearly
greater than for other TLAAS-eligible manufacturers. Compared with
other TLAAS-eligible manufacturers, these companies also have more
limited fleets across which to average the standards. Some companies
have only a few vehicle models all of a similar utility, and thus their
averaging abilities are extremely limited posing lead time issues of
greater severity than other TLAAS-eligible manufacturers. EPA's
feasibility analysis provided in Section III.D., shows that these
companies face a compliance shortfall significantly greater than other
TLAAS companies (see Table III.D.6-4). This shortfall is primarily due
to their narrow product lines and more limited ability to average
across their vehicle fleets. In addition, with fewer models with which
to average, there is a higher likelihood that phase-in requirements may
conflict with normal product redesign cycles.
Therefore, for manufacturers with MY 2009 U.S. sales of less than
50,000 vehicles, EPA is finalizing additional TLAAS compliance
flexibility through model year 2016. These manufacturers will be
allowed to place up to 200,000 vehicles in the TLAAS program in MY
2012-2015 and an additional 50,000 vehicles in MY 2016. To be eligible
for the additional allotment above the base TLAAS level of 100,000
vehicles, manufacturers must annually demonstrate that they have
diligently made a good faith effort to purchase credits from other
manufacturers in order to comply with the base TLAAS program, but that
sufficient credits were not available. Manufacturers must secure
credits to the extent they are reasonably available from other
manufacturers to offset the difference between their emissions
reductions obligations under the base TLAAS program and the expanded
TLAAS program. Manufacturers must document their efforts to purchase
credits as part of their end of year compliance report. All other
aspects of the TLAAS program including the 1.25x adjustment to the
standards and the credits provision restrictions remain the same as
described above for the same reasons. This will still require the
manufacturers to reduce emissions significantly in the 2012-2016 time-
frame and to meet the final emissions standards in MY 2017. The
standards remain very challenging for these manufacturers but these
additional provisions will allow them the necessary lead time for
implementing their strategy for compliance with the final, most
stringent standards.
The eligibility limit of 50,000 vehicles will be treated in a
similar way as the 400,000 vehicle eligibility limit is treated, as
described above. Manufacturers with model year 2009 U.S. sales of less
than 50,000 vehicles are eligible for the expanded TLAAS flexibility.
Manufacturers whose sales grow in later years above 50,000 vehicles
without merger or acquisition will continue to be eligible for the
expanded TLAAS program. However, manufacturers that exceed the 50,000
vehicle limit through mergers or acquisitions will not be eligible for
the expanded TLAAS program in the model year following the merger or
acquisition, but may continue to be eligible for the base TLAAS program
if the MY 2009 sales of the new company would have been below the
400,000 vehicle eligibility cut point. The use of TLAAS by all the
entities within the company in years prior to the merger must be
counted against the 100,000 vehicle limit of the base program. If the
100,000 vehicle limit has been exceeded, the company is no longer
eligible for TLAAS.
6. Deferment of CO2 Standards for Small Volume Manufacturers
With Annual Sales Less Than 5,000 Vehicles
In the proposal, in the context of the TLAAS program, EPA
recognized that there would be a wide range of companies within the
eligible manufacturers with sales less than 400,000 vehicles in model
year 2009. As noted in the proposal, some of these companies, while
having relatively small U.S. sales volumes, are large global automotive
firms, including companies such as Mercedes and Volkswagen. Other
companies are significantly smaller niche firms, with sales volumes
closer to 10,000 vehicles per year worldwide, such as Aston Martin. EPA
anticipated that there is a small number of such smaller volume
manufacturers, which may face greater challenges in meeting the
standards due to their limited product lines across which to average.
EPA requested comment on whether the proposed TLAAS program would
provide sufficient lead-time for these smaller firms to incorporate the
technology needed to comply with the proposed GHG standards. See 74 FR
at 49524.

[[Page 25420]]

EPA received comments from several very small volume manufacturers
that the TLAAS program would not provide sufficient lead time, as
described above. EPA agrees with comments that the standards would be
extremely challenging and potentially infeasible for these small volume
manufacturers, absent credits from other manufacturers, and that credit
availability at this point is highly uncertain--although these
companies are planning to introduce significant GHG-reducing
technologies to their product lines, they are still highly unlikely to
meet the standards by MY 2016. Because the products produced by these
manufacturers are so unique, these manufacturers were not included in
EPA's OMEGA modeling assessment of the technology feasibility and costs
to meet the proposed standards. As noted above, these manufacturers
have only a few models and have very high baseline emissions. TLAAS
manufacturers are projected to be required to reduce emissions by up to
39%, whereas SVMs in many cases would need to cut their emissions by
more than half to comply with MY 2016 standards.
Given the unique feasibility issues raised for these manufacturers,
EPA is deferring establishing CO2 standards for
manufacturers with U.S. sales of less than 5,000 vehicles.\188\ This
will provide EPA more time to consider the unique challenges faced by
these manufacturers. EPA expects to conduct this rulemaking in the 2012
timeframe. The deferment only applies to CO2 standards and
SVMs must meet N2O and CH4 standards. EPA plans
to set standards for these manufacturers as part of a future rulemaking
in the next 18 months. This future rulemaking will allow EPA to fully
examine the technologies and emissions levels of vehicles offered by
small manufacturers and to determine the potential emissions control
capabilities, costs, and necessary lead time. This timing may also
allow a credits market to develop, so that EPA may consider the
availability of credits during the rulemaking process. See State of
Mass. v. EPA, 549 U.S. at 533 (EPA retains discretion as to timing of
any regulations addressing vehicular GHG emissions under section
202(a)(1)). We expect that standards would begin to be implemented in
the MY 2016 timeframe. This approach is consistent with that envisioned
by California for these manufacturers. EPA estimates that eligible
small volume manufacturers currently comprise less than 0.1 percent of
the total light-duty vehicle sales in the U.S., and therefore the
deferment will have a very small impact on the GHG emissions reductions
from the standards.
---------------------------------------------------------------------------

\188\ See final regulations at 40 CFR 86.1801-12(k).
---------------------------------------------------------------------------

In addition to the 5,000 vehicle per year cut point, to be eligible
for deferment each year, manufacturers must also demonstrate due
diligence in attempting to secure credits from other manufacturers.
Manufacturers must make a good faith effort to secure credits to the
extent they are reasonably available from other manufacturers to offset
the difference between their baseline emissions and what their
obligations would be under the TLAAS program starting in MY 2012.
Eligibility will be determined somewhat differently compared to the
TLAAS program. Manufacturers with either MY 2008 or MY 2009 U.S. sales
of less than 5,000 vehicles will be initially eligible. This includes
``branded sales'' for companies that sold vehicles under a larger
manufacturer but has become independent by the end of calendar year
2010. EPA is including MY 2008 as well as MY 2009 because some
manufacturers in this market segment have such limited sales that they
often drop in and out of the market from year to year.
In determining eligibility, manufacturers must be aggregated
according to the provisions of 40 CFR 86.1838-01(b)(3), which requires
the sales of different firms to be aggregated in various situations,
including where one firm has a 10% or more equity ownership of another
firm, or where a third party has a 10% or more equity ownership of two
or more firms. EPA received public comment from a manufacturer
requesting that EPA should allow a manufacturer to apply to EPA to
establish small volume manufacturer status based on the independence of
its research, development, testing, design, and manufacturing from
another firm that may have an ownership interest in that manufacturer.
EPA has reviewed this comment, but is not finalizing such a provision
at this time. EPA believes that this issue likely presents some
competitive issues, which we would like to be fully considered through
the public comment process. Therefore, EPA plans to consider this issue
and seek public comments in our proposal for small volume manufacturer
CO2 standards, which we expect to complete within 18 months.
To remain eligible for the deferral from standards, the rolling
average of three consecutive model years of sales must remain below
5,000 vehicles. EPA is establishing the 5,000 vehicle threshold to
allow for some sales growth by SVMs, as SVMs typically have annual
sales of below 2,000 vehicles. However, EPA wants to ensure that
standards for as few vehicles as possible are deferred and therefore
believes it is appropriate that manufacturers with U.S. sales growing
to above 5,000 vehicles per year be required to comply with standards
(including TLAAS, as applicable). Manufacturers with unusually strong
sales in a given year would still likely remain eligible, based on the
three year rolling average. However, if a manufacturer takes steps to
expand in the U.S. market on a permanent basis such that they
consistently sell more than 5,000 vehicles per year, they must meet the
TLAAS standards. EPA believes a manufacturer will be able to consider
these provisions, along with other factors, in its planning to
significantly expand in the U.S. market.
For manufacturers exceeding the 5,000 vehicle rolling average
through mergers or acquisitions of other manufacturers, those
manufacturers will lose eligibility in the MY immediately following the
last year of the rolling average. For manufacturers exceeding this
level through sales growth, but remaining below a 50,000 vehicle
threshold, the manufacturer will lose eligibility for the deferred
standards in the second model year following the last year of the
rolling average. For example, if the rolling average of MYs 2009-2011
exceeded 5,000 vehicles but was below 50,000 vehicles, the manufacturer
would not be eligible for the deferred standards in MY 2013. For
manufacturers with a 3-year rolling average exceeding 50,000 vehicles,
the manufacturer would lose eligibility in the MY immediately following
the last model year in the rolling average. For example, if the rolling
average of MYs 2009-2011 exceeded 50,000 vehicles, the manufacturer
would not be eligible for the deferred standards in MY 2012. Such
manufacturers may continue to be eligible for TLAAS, or the expanded
TLAAS program, per the provisions described above. EPA believes these
provisions are needed to ensure that the SVM deferment remains targeted
to true small volume manufacturers and does not become available to
larger manufacturers through mergers or acquisitions. EPA is including
the 50,000 vehicle criteria to differentiate between manufacturers that
may slowly gain more sales and manufacturers that have taken major
steps to significantly increase their presence in the U.S. market, such
as by introducing new vehicle models. EPA believes manufacturers
selling more than 50,000

[[Page 25421]]

vehicles should not be able to take advantage of the deferment, as they
should be able to meet the applicable TLAAS standards through averaging
across their larger product line.
EPA is requiring that potential SVMs submit a declaration to EPA
containing a detailed written description of how the manufacturer
qualifies as a small volume manufacturer. The declaration must contain
eligibility information including MY 2008 and 2009 U.S. sales, the last
three completed MYs sales information, detailed information regarding
ownership relationships with other manufacturers, and documentation of
efforts to purchase credits from other manufacturers. Because such
manufacturers are not automatically exempted from other EPA regulations
for light-duty vehicles and light-duty trucks, entities are subject to
the greenhouse gas control requirements in this program until such a
declaration has been submitted and approved by EPA. The declaration
must be submitted annually at the time of vehicle emissions
certification under the EPA Tier 2 program, beginning in MY 2012.
7. Nitrous Oxide and Methane Standards
In addition to fleet-average CO2 standards, as proposed,
EPA is establishing separate per-vehicle standards for nitrous oxide
(N2O) and methane (CH4) emissions.\189\ The
agency's intention is to set emissions standards that act to cap
emissions to ensure that future vehicles do not increase their
N2O and CH4 emissions above levels typical of
today's vehicles. EPA proposed to cap N2O at a level of
0.010 g/mi and to cap CH4 at a level of 0.03 g/mi. Both of
these compounds are more potent contributors to global warming than
CO2; N2O has a global warming potential, or GWP,
of 298 and CH4 has a GWP of 25.\190\
---------------------------------------------------------------------------

\189\ See final regulations at 40 CFR 86.1818-12(f).
\190\ The global warming potentials (GWP) used in this rule are
consistent with the Intergovernmental Panel on Climate Change (IPCC)
Fourth Assessment Report (AR4).
---------------------------------------------------------------------------

EPA received many comments on the proposed N2O and
CH4 standards. A range of stakeholders supported the
proposed approach of ``cap'' standards and the proposed emission
levels, including most states and environmental organizations that
addressed this topic, and the Manufacturers of Emissions Control
Association. These commenters stated that EPA needs to address all
mobile GHGs under the Clean Air Act, and N2O and
CH4 are both more potent contributors to global warming than
CO2. The Center for Biological Diversity commented that in
light of the potency of these GHGs, EPA should develop standards which
reduce emissions over current levels and that EPA had not analyzed
either the technologies or the costs of doing so. EPA discusses these
comments and our responses below and in the Response to Comments
Document.
Auto manufacturers generally did not support standards for these
GHGs, stating that the levels of these GHGs from current vehicles are
too small to warrant standards at this time. These commenters also
stated that if EPA were to proceed with ``cap'' standards, the
stringency of the proposed levels could restrict the introduction of
some new technologies. Commenters specifically raised this concern with
the examples of diesel and lean-burn gasoline for N2O, or
natural gas and ethanol fueled vehicles for CH4. Only one
manufacturer, Volkswagen, submitted actual test data to support these
claims; very limited emission data on two concept vehicles--a CNG
vehicle and a flexible-fuel vehicle--indicated measured emission levels
near or above the proposed standards, but included no indication of
whether any technological steps had been taken to reduce emissions
below the cap levels. Many commenters support an approach of
establishing a CO2-equivalent standard, where N2O
and CH4 could be averaged with CO2 emissions to
result in an overall CO2-equivalent compliance value,
similar to the approach California has used for its GHG standards \191\
Under such an approach, the auto industry commenters supported using a
default value for N2O emissions in lieu of a measured test
value. Several auto manufacturers also had concerns that a new
requirement to measure N2O would require significant
equipment and facility upgrades and would create testing challenges
with new measurement equipment with which they have little experience.
---------------------------------------------------------------------------

\191\ California Environmental Protection Agency Air Resources
Board, Staff Report: Initial Statement of Reasons for Proposed
Rulemaking Public Hearing To Consider Adoption of Regulations To
Control Greenhouse Gas Emissions From Motor Vehicles, August 6,
2004.
---------------------------------------------------------------------------

EPA has considered these comments and is finalizing the cap
standards for N2O and CH4 as proposed. EPA agrees
with the NGO, State, and other commenters that light-duty vehicle
emissions are small but important contributors to the U.S.
N2O and CH4 inventories, and that in the absence
of a limitation, the potential for significant emission increases
exists with the evolution of new vehicle and engine technologies.
(Indeed, the industry commenters concede as much in stating that they
are contemplating introducing vehicle technologies that could result in
emissions exceeding the cap standard levels). EPA also believes that in
most cases N2O and CH4 emissions from light-duty
vehicles will remain well below the cap standards. Therefore, we are
setting cap standards for these GHGs at the proposed levels. However,
as described below, the agency is incorporating several provisions
intended to address industry concerns about technological feasibility
and leadtime, including an optional CO2-equivalent approach
and, for N2O, more leadtime before testing will be required
to demonstrate compliance with the emissions standard (in interim,
manufacturers may certify based on a compliance statement based on good
engineering judgment).
a. Nitrous Oxide (N2O) Exhaust Emission Standard
As stated above, N2O is a global warming gas with a high
global warming potential.\192\ It accounts for about 2.3% of the
current greenhouse gas emissions from cars and light trucks.\193\ EPA
is setting a per-vehicle N2O emission standard of 0.010 g/
mi, measured over the traditional FTP vehicle laboratory test cycles.
The standard will become effective in model year 2012 for all light-
duty cars and trucks. The standard is designed to prevent increases in
N2O emissions from current levels; i.e., it is a no-
backsliding standard.
---------------------------------------------------------------------------

\192\ N2O has a GWP of 298 according to the IPCC
Fourth Assessment Report (AR4).
\193\ See RIA Chapter 2.
---------------------------------------------------------------------------

N2O is emitted from gasoline and diesel vehicles mainly
during specific catalyst temperature conditions conducive to
N2O formation. Specifically, N2O can be generated
during periods of emission hardware warm-up when rising catalyst
temperatures pass through the temperature window when N2O
formation potential is possible. For current Tier 2 compatible gasoline
engines with conventional three-way catalyst technology, N2O
is not generally produced in significant amounts because the time the
catalyst spends at the critical temperatures during warm-up is short.
This is largely due to the need to quickly reach the higher
temperatures necessary for high catalyst efficiency to achieve emission
compliance for criteria pollutants. As several auto manufacturer
comments noted, N2O is a more significant concern with
diesel vehicles, and potentially future gasoline lean-burn engines,
equipped with advanced catalytic NOX

[[Page 25422]]

emissions control systems. In the absence of N2O emission
standards, these systems could be designed in a way that emphasizes
efficient NOX control while at the same time allowing the
formation of significant quantities of N2O. Excess oxygen
present in the exhaust during lean-burn conditions in diesel or lean-
burn gasoline engines equipped with these advanced systems can favor
N2O formation if catalyst temperatures are not carefully
controlled. Without specific attention to controlling N2O
emissions in the development of such new NOX control
systems, vehicles could have N2O emissions many times
greater than are emitted by current gasoline vehicles.
EPA is setting an N2O emission standard that the agency
believes will be met by current-technology gasoline vehicles at
essentially no cost. As just noted, N2O formation in current
catalyst systems occurs, but the emission levels are relatively low,
because the time the catalyst spends at the critical temperatures
during warm-up when N2O can form is short. At the same time,
EPA believes that the standard will ensure that the design of advanced
NOX control systems, especially for future diesel and lean-
burn gasoline vehicles, will control N2O emission levels.
While current NOX control approaches used on current Tier 2
diesel vehicles do not tend to favor the formation of N2O
emissions, EPA believes that this N2O standard will
discourage new emission control designs that achieve criteria emissions
compliance at the cost of increased N2O emissions. Thus, the
standard will cap N2O emission levels, with the expectation
that current gasoline and diesel vehicle control approaches that comply
with the Tier 2 vehicle emission standards for NOX will not
increase their emission levels, and that the cap will ensure that
future vehicle designs will be appropriately controlled for
N2O emissions.
The level of the N2O standard is approximately two times
the average N2O level of current gasoline passenger cars and
light-duty trucks that meet the Tier 2 NOX standards. EPA
has not previously regulated N2O emissions, and available
data on current vehicles is limited. However, EPA derived the standard
from a combination of emission factor values used in modeling light
duty vehicle emissions and limited recent EPA test
data.^{194 195} Because the standard represents a level 100
percent higher than the average current N2O level, we
continue to believe that most if not all Tier 2 compliant gasoline and
diesel vehicles will easily be able to meet the standards.
Manufacturers typically use design targets for NOX emission
levels of about 50% of the standard, to account for in-use emissions
deterioration and normal testing and production variability, and EPA
expects that manufacturers will use a similar approach for
N2O emission compliance. EPA did not propose and is not
finalizing a more stringent standard for current vehicles because we
believe that the stringent Tier 2 program and the associated
NOX fleet average requirement already result in significant
N2O control, and the agency does not expect current
N2O levels to rise for these vehicles. Moreover, EPA
believes that the CO2 standards will be challenging for the
industry and that these standards should be the industry's chief focus
in this first phase of vehicular GHG emission controls. See
Massachusetts v. EPA, 549 U.S. at 533 (EPA has significant discretion
as to timing of GHG regulations); see also Sierra Club v. EPA, 325 F.
3d 374, 379 (DC Cir. 2003) (upholding anti-backsliding standards for
air toxics under technology-forcing section 202 (l) because it is
reasonable for EPA to assess the effects of its other regulations on
the motor vehicle sector before aggressively regulating emissions of
toxic vehicular air pollutants.
---------------------------------------------------------------------------

\194\ Memo to docket ``Derivation of Proposed N2O and
CH4 Cap Standards,'' Tad Wysor, EPA, November 19, 2009.
Docket EPA-HQ-OAR-2009-0472-6801.
\195\ Memo to docket ``EPA NVFEL N2O Test Data,''
Tony Fernandez, EPA.
---------------------------------------------------------------------------

Diesel cars and light trucks with advanced emission control
technology are in the early stages of development and
commercialization. As this segment of the vehicle market develops, the
N2O standard will likely require these manufacturers to
incorporate control strategies that minimize N2O formation.
Available approaches include using electronic controls to limit
catalyst conditions that might favor N2O formation and
consider different catalyst formulations. While some of these
approaches may have modest associated costs, EPA believes that they
will be small compared to the overall costs of the advanced
NOX control technologies already required to meet Tier 2
standards.
In the proposal, EPA sought comment on an approach of expressing
N2O and CH4 in common terms of CO2-
equivalent emissions and combining them into a single standard along
with CO2 emissions. 74 FR at 49524. California's ``Pavley''
program adopted such a CO2-equivalent emissions standards
approach to GHG emissions.\196\ EPA was primarily concerned that such
an approach could undermine the stringency of the CO2
standards, as the proposed standards were designed to ``cap''
N2O and CH4 emissions, rather than reflecting a
level either that is the industry fleet-wide average or that would
effect reductions in these GHGs.
---------------------------------------------------------------------------

\196\ California Environmental Protection Agency Air Resources
Board, Staff Report: Initial Statement of Reasons for Proposed
Rulemaking Public Hearing To Consider Adoption of Regulations To
Control Greenhouse Gas Emissions From Motor Vehicles, August 6,
2004.
---------------------------------------------------------------------------

As noted above, several auto manufacturers expressed interest in
such a CO2-equivalent approach, due to concerns that the
caps could be limiting for some advanced technology vehicles. While we
continue to believe that the vast majority of light-duty vehicles will
be able to easily meet the standards, we acknowledge that advanced
diesel or lean-burn gasoline vehicles of the future may face slightly
greater challenges. Therefore, after considering these comments, EPA is
finalizing an optional compliance approach to provide flexibility for
any advanced technologies that may have challenges in meeting the
N2O or CH4 cap standards.
In lieu of complying with the separate N2O and
CH4 cap standards, a manufacturer may choose to comply with
a CO2-equivalent standard. A manufacturer choosing this
option will convert its N2O and CH4 test results
(or, as described below, a default N2O value for MY 2012-
2014) into CO2-equivalent values and add this sum to their
CO2 emissions. This CO2-equivalent value will
still need to comply with the manufacturer's footprint-based
CO2 target level. In other words, a manufacturer could
offset any N2O emissions (or any CH4 emissions)
by taking steps to further reduce CO2. A manufacturer
choosing this option will need to apply this approach to all of the
test groups in its fleet. This approach is more environmentally
protective overall than the cap standard approach, since the
manufacturer will need to reduce its CO2 emissions to offset
the higher N2O (or CH4) levels, but will not be
allowed to increase CO2 above its footprint target level by
reducing N2O (or CH4).
The compliance level in g/mi for the optional CO2-
equivalent approach for gasoline vehicles is calculated as
CO2 + (CWF/0.273 x NMHC) + (1.571 x CO) + (298 x
N2O) + (25 x CH4).\197\ The N2O and
CH4 values are the measured emission values for these GHGs,
except N2O in model years 2012 through 2014. For these model
years, manufacturers may use a default N2O value of 0.010

[[Page 25423]]

g/mi, the same value as the N2O cap standard. For MY 2015
and later, the manufacturer would need to provide actual test data on
the emission data vehicle for each test group. (That is, N2O
data would not be required for each model type, since EPA believes that
there will likely be little N2O variability among model
types within a test group.) EPA believes that its selection of 0.010 g/
mi as the N2O default value is an appropriately protective
level, on the high end of current technologies, as further discussed
below. Consistent with the other elements of the equation,
N2O and CH4 must be included at full useful life
deteriorated values. This requires testing using the highway test cycle
in addition to the FTP during the manufacturer's deterioration factor
(DF) development program. However, EPA recognizes that manufacturers
may not be able to develop DFs for N2O and CH4
for all their vehicles in the 2012 model year, and thus EPA is allowing
the use of alternative values through the 2014 model year. For
N2O the alternative value is the DF developed for
NOX emissions, and for CH4 the alternative value
is the DF developed for NMOG emissions. Finally, for manufacturers
using this option, the CO2-equivalent emission level would
also be the basis for any credits that the manufacturer might generate.
---------------------------------------------------------------------------

\197\ This equation will differ depending upon the fuel; see the
final regulations for equations for other fuels.
---------------------------------------------------------------------------

Manufacturers expressed concerns about their ability to acquire and
install N2O analytical equipment. However, the agency
continues to believe that such burdens, while not trivial, will also
not be excessive. While many manufacturers do not appear to have
invested yet in adding N2O measurement equipment to their
test facilities, EPA is not aware of any information to indicate that
that suppliers will have difficulty providing sufficient hardware, or
that such equipment is unusually expensive or complex compared to
existing measurement hardware. EPA allows N2O measurement
using any of four methods, all of which are commercially available
today. The costs of certification and other indirect costs of this rule
are accounted for in the Indirect Cost Multipliers, discussed in
Section III.H below.
Still, given the short lead-time for this rule and the newness of
N2O testing to this industry, EPA proposed that
manufacturers be able to apply for a certificate of conformity with the
N2O standard for model year 2012 provided that they supply a
compliance statement based on good engineering judgment. Under the
proposal, beginning in MY 2013, manufacturers would have needed to base
certification on actual N2O testing data. This approach was
intended to reasonably ensure that the emission standards are being
met, while allowing manufacturers lead-time to purchase new
N2O emissions measurement equipment, modify certification
test facilities, and begin N2O testing. After consideration
of the comments, EPA agrees with manufacturers that one year of
additional lead-time to begin actual N2O measurement across
their vehicle fleets may still be insufficient for manufacturers to
efficiently make the necessary facility changes and equipment
purchases. Therefore, EPA is extending the ability to certify based on
a compliance statement for two additional years, through model year
2014. For 2015 and later model years, manufacturers will need to submit
measurements of N2O for compliance purposes.
b. Methane (CH4) Exhaust Emission Standard
Methane (CH4) is a greenhouse gas with a high global
warming potential.\198\ It accounts for about 0.2% of the greenhouse
gases from cars and light trucks.\199\
---------------------------------------------------------------------------

\198\ CH4 has a GWP of 25 according to the IPCC
Fourth Assessment Report (AR4).
\199\ See RIA Chapter 2.
---------------------------------------------------------------------------

EPA is setting a CH4 emission standard of 0.030 g/mi as
measured on the FTP, to apply beginning with model year 2012 for both
cars and trucks. EPA believes that this level for the standard will be
met by current gasoline and diesel vehicles, and will prevent large
increases in future CH4 emissions. This is particularly a
concern in the event that alternative fueled vehicles with high methane
emissions, like some past dedicated compressed natural gas (CNG)
vehicles and some flexible-fueled vehicles when operated on E85 fuel,
become a significant part of the vehicle fleet. Currently EPA does not
have separate CH4 standards because unlike other
hydrocarbons it does not contribute significantly to ozone
formation.\200\ However, CH4 emissions levels in the
gasoline and diesel car and light truck fleet have nevertheless
generally been controlled by the Tier 2 standards for non-methane
organic gases (NMOG). However, without an emission standard for
CH4, there is no guarantee that future emission levels of
CH4 will remain at current levels as vehicle technologies
and fuels evolve.
---------------------------------------------------------------------------

\200\ But see Ford Motor Co. v. EPA, 604 F. 2d 685 (D.C. Cir.
1979) (permissible for EPA to regulate CH4 under CAA
section 202(b)).
---------------------------------------------------------------------------

The standard will cap CH4 emission levels, with the
expectation that emissions levels of current gasoline and diesel
vehicles meeting the Tier 2 emission standards will not increase. The
level of the standard will generally be achievable for typical vehicles
through normal emission control methods already required to meet the
Tier 2 emission standards for NMOG. Also, since CH4 is
already measured under the current Tier 2 regulations (so that it may
be subtracted to calculate non-methane hydrocarbons), we believe that
the standard will not result in any additional testing costs.
Therefore, EPA is not attributing any costs to this part of this
program. Since CH4 is produced during fuel combustion in
gasoline and diesel engines similarly to other hydrocarbon components,
controls targeted at reducing overall NMOG levels are generally also
effective in reducing CH4 emissions. Therefore, for typical
gasoline and diesel vehicles, manufacturer strategies to comply with
the Tier 2 NMOG standards have to date tended to prevent increases in
CH4 emissions levels. The CH4 standard will
ensure that emissions will be addressed if in the future there are
increases in the use of natural gas or other alternative fuels or
technologies that may result in higher CH4 emissions.
As with the N2O standard, EPA is setting the level of
the CH4 standard to be approximately two times the level of
average CH4 emissions from Tier 2 gasoline passenger cars
and light-duty trucks. EPA believes the standard will easily be met by
current gasoline vehicles, and that flexible fuel vehicles operating on
ethanol can be designed to resolve any potential CH4
emissions concerns. Similarly, since current diesel vehicles generally
have even lower CH4 emissions than gasoline vehicles, EPA
believes that diesels will also meet the CH4 standard.
However, EPA also believes that to set a CH4 emission
standard more stringent than the proposed standard could effectively
make the Tier 2 NMOG standard more stringent and is inappropriate for
that reason (and untimely as well, given the challenge of meeting the
CO2 standards, as noted above).
Some CNG-fueled vehicles have historically produced significantly
higher CH4 emissions than gasoline or diesel vehicles. This
is because CNG fuel is essentially methane and any unburned fuel that
escapes combustion and is not oxidized by the catalyst is emitted as
methane. However, in recent model years, the few dedicated CNG vehicles
sold in the U.S. meeting the Tier 2 standards have had CH4
control as effective as that of gasoline or diesel vehicles. Still,
even if these vehicles meet the Tier 2 NMOG standard and appear to have
effective CH4 control by

[[Page 25424]]

nature of the NMOG controls, Tier 2 standards do not require
CH4 control. Although EPA believes that in most cases that
the CH4 cap standard should not require any different
emission control designs beyond what is already required to meet Tier 2
NMOG standards on a dedicated CNG vehicle, the cap will ensure that
systems maintain the current level of CH4 control.
Some manufacturers have also expressed some concerns about
CH4 emissions from flexible-fueled vehicles operating on E85
(85% ethanol, 15% gasoline). However, we are not aware of any
information that would indicate that if engine-out CH4
proves to be higher than for a typical gasoline vehicle, that such
emissions could not be managed by reasonably available control
strategies (perhaps similar to those used in dedicated CNG vehicles).
As described above, in response to the comments, EPA will also
allow manufacturers to choose to comply with a CO2-
equivalent standard in lieu of complying with a separate CH4
cap standard. A manufacturer choosing this option would convert its
N2O and CH4 test results into CO2-
equivalent values (using the respective GWP values), and would then
compare this value to the manufacturer's footprint-based CO2
target level to determine compliance. However, as with N2O,
this approach will not permit a manufacturer to increase its
CO2 by reducing CH4; the company's footprint-
based CO2 target level would remain the same.
8. Small Entity Exemption
As proposed, EPA is exempting from GHG emissions standards small
entities meeting the Small Business Administration (SBA) size criteria
of a small business as described in 13 CFR 121.201.\201\ EPA will
instead consider appropriate GHG standards for these entities as part
of a future regulatory action. This includes both U.S.-based and
foreign small entities in three distinct categories of businesses for
light-duty vehicles: small volume manufacturers, independent commercial
importers (ICIs), and alternative fuel vehicle converters.
---------------------------------------------------------------------------

\201\ See final regulations at 40 CFR 86.1801-12(j).
---------------------------------------------------------------------------

EPA has identified about 13 entities that fit the Small Business
Administration (SBA) size criterion of a small business. EPA estimates
there currently are approximately two small volume manufacturers, eight
ICIs, and three alternative fuel vehicle converters in the light-duty
vehicle market. Further detail is provided in Section III.I.3, below.
EPA estimates that these small entities comprise less than 0.1 percent
of the total light-duty vehicle sales in the U.S., and therefore the
exemption will have a negligible impact on the GHG emissions reductions
from the standards.
To ensure that EPA is aware of which companies would be exempt, EPA
proposed to require that such entities submit a declaration to EPA
containing a detailed written description of how that manufacturer
qualifies as a small entity under the provisions of 13 CFR 121.201. EPA
has reconsidered the need for this additional submission under the
regulations and is deleting it as not necessary. We already have
information on the limited number of small entities that we expect
would receive the benefits of the exemption, and do not need the
proposed regulatory requirement to be able to effectively implement
this exemption for those parties who in fact meet its terms. Small
entities are currently covered by a number of EPA motor vehicle
emission regulations, and they routinely submit information and data on
an annual basis as part of their compliance responsibilities.
EPA did not receive adverse comments regarding the proposed small
entity exemption. EPA received comments concerning whether or not the
small entity exemption applies to foreign manufacturers. EPA clarifies
that foreign manufacturers meeting the SBA size criteria are eligible
for the exemption, as was EPA's intent during the proposal.

C. Additional Credit Opportunities for CO2 Fleet Average Program

The final standards represent a significant multi-year challenge
for manufacturers, especially in the early years of the program.
Section III.B.4 above describes EPA's provisions for manufacturers to
be able to generate credits by achieving fleet average CO2
emissions below their fleet average standard, and also how
manufacturers can use credits to comply with the standards. As
described in Section III.B.4, credits can be carried forward five
years, carried back three years, transferred between vehicle
categories, and traded between manufacturers. The credits provisions
described below provide manufacturers with additional ways to earn
credits starting in MY 2012. EPA is also including early credits
provisions for the 2009-2011 model years, as described below in Section
III.C.5.
The provisions described below provide additional flexibility,
especially in the early years of the program. This helps to address
issues of lead-time or technical feasibility for various manufacturers
and in several cases provides an incentive for promotion of technology
pathways that warrant further development. EPA is finalizing a variety
of credit opportunities because manufacturers are not likely to be in a
position to use every credit provision. EPA expects that manufacturers
are likely to select the credit opportunities that best fit their
future plans.
EPA believes it is critical that manufacturers have options to ease
the transition to the final MY 2016 standards. At the same time, EPA
believes these credit programs must be and are designed in a way to
ensure that they achieve emission reductions that achieve real-world
reductions over the full useful life of the vehicle (or, in the case of
FFV credits and Advanced Technology incentives, to incentivize the
introduction of those vehicle technologies) and are verifiable. In
addition, EPA believes that these credit programs do not provide an
opportunity for manufacturers to earn ``windfall'' credits. Comments on
the proposed EPA credit programs are summarized below along with EPA's
response, and are detailed in the Response to Comments document.
1. Air Conditioning Related Credits
Manufacturers will be able to generate and use credits for improved
air conditioner (A/C) systems in complying with the CO2
fleetwide average standards described above (or otherwise to be able to
bank or trade the credits). EPA expects that most manufacturers will
choose to utilize the A/C provisions as part of its compliance
demonstration (and for this reason cost of compliance with A/C related
emission reductions are assumed in the cost analysis). The A/C
provisions are structured as credits, unlike the CO2
standards for which manufacturers will demonstrate compliance using 2-
cycle (city/highway) tests (see Sections III.B and III.E.). Those tests
do not measure either A/C leakage or tailpipe CO2 emissions
attributable to A/C load. Thus, it is a manufacturer's option to
include A/C GHG emission reductions as an aspect of its compliance
demonstration. Since this is an elective alternative, EPA is referring
to the A/C part of the rule as a credit.
EPA estimates that direct A/C GHG emissions--emissions due to the
leakage of the hydrofluorocarbon refrigerant in common use today--
account for 5.1% of CO2-equivalent GHGs from light-duty cars
and trucks. This includes the direct leakage of refrigerant as well as
the subsequent leakage associated with maintenance and servicing, and
with disposal at the end of the vehicle's life.

[[Page 25425]]

The emissions that are associated with leakage reductions are the
direct leakage and the leakage associated with maintenance and
servicing. Together these are equivalent to CO2 emissions of
approximately 13.6 g/mi per car and light-truck. EPA also estimates
that indirect GHG emissions (additional CO2 emitted due to
the load of the A/C system on the engine) account for another 3.9% of
light-duty GHG emissions.\202\ This is equivalent to CO2
emissions of approximately 14.2 g/mi per vehicle. The derivation of
these figures can be found in Chapter 2.2 of the EPA RIA.
---------------------------------------------------------------------------

\202\ See Chapter 2, Section 2.2.1.2 of the RIA.
---------------------------------------------------------------------------

EPA believes that it is important to address A/C direct and
indirect emissions because the technologies that manufacturers will
employ to reduce vehicle exhaust CO2 will have little or no
impact on A/C related emissions. Without addressing A/C related
emissions, as vehicles become more efficient, the A/C related
contribution will become a much larger portion of the overall vehicle
GHG emissions.
Over 95% of the new cars and light trucks in the United States are
equipped with A/C systems and, as noted, there are two mechanisms by
which A/C systems contribute to the emissions of greenhouse gases:
Through leakage of refrigerant into the atmosphere and through the
consumption of fuel to provide mechanical power to the A/C system. With
leakage, it is the high global warming potential (GWP) of the current
automotive refrigerant (HFC-134a, with a GWP of 1430) that results in
the CO2-equivalent impact of 13.6 g/mi.\203\ Due to the high
GWP of this HFC, a small leakage of the refrigerant has a much greater
global warming impact than a similar amount of emissions of
CO2 or other mobile source GHGs. Manufacturers can reduce A/
C leakage emissions by using leak-tight components. Also, manufacturers
can largely eliminate the global warming impact of leakage emissions by
adopting systems that use an alternative, low-GWP refrigerant, as
discussed below.\204\ The A/C system also contributes to increased
CO2 emissions through the additional work required to
operate the compressor, fans, and blowers. This additional work
typically is provided through the engine's crankshaft, and delivered
via belt drive to the alternator (which provides electric energy for
powering the fans and blowers) and the A/C compressor (which
pressurizes the refrigerant during A/C operation). The additional fuel
used to supply the power through the crankshaft necessary to operate
the A/C system is converted into CO2 by the engine during
combustion. This incremental CO2 produced from A/C operation
can thus be reduced by increasing the overall efficiency of the
vehicle's A/C system, which in turn will reduce the additional load on
the engine from A/C operation.\205\
---------------------------------------------------------------------------

\203\ The global warming potentials (GWP) used in this rule are
consistent with Intergovernmental Panel on Climate Change (IPCC)
Fourth Assessment Report (AR4). (At this time, the IPCC Second
Assessment Report (SAR) GWP values are used in the official U.S.
greenhouse gas inventory submission to the climate change
framework.)
\204\ Refrigerant emissions during maintenance and at the end of
the vehicle's life (as well as emissions during the initial charging
of the system with refrigerant) are also addressed by the CAA Title
VI stratospheric ozone program, as described below.
\205\ We chose not to address changes to the weight of the A/C
system, since the issue of CO2 emissions from the fuel
consumption of normal (non-A/C) operation, including basic vehicle
weight, is inherently addressed by the primary CO2
standards (Section III.B above).
---------------------------------------------------------------------------

Manufacturers can make very feasible improvements to their A/C
systems to address A/C system leakage and efficiency. EPA is finalizing
two separate credit approaches to address leakage reductions and
efficiency improvements independently. A leakage reduction credit will
take into account the various technologies that could be used to reduce
the GHG impact of refrigerant leakage, including the use of an
alternative refrigerant with a lower GWP. An efficiency improvement
credit will account for the various types of hardware and control of
that hardware available to increase the A/C system efficiency. For
purposes of use of A/C credits at certification, manufacturers will be
required to attest to the durability of the leakage reduction and the
efficiency improvement technologies over the full useful life of the
vehicle.
EPA believes that both reducing A/C system leakage and increasing
efficiency are highly cost-effective and technologically feasible. EPA
expects most manufacturers will choose to use these A/C credit
provisions, although some may not find it necessary to do so.
a. A/C Leakage Credits
The refrigerant used in vehicle A/C systems can get into the
atmosphere by many different means. These refrigerant emissions occur
from the slow leakage over time that all closed high pressure systems
will experience. Refrigerant loss occurs from permeation through hoses
and leakage at connectors and other parts where the containment of the
system is compromised. The rate of leakage can increase due to
deterioration of parts and connections as well. In addition, there are
emissions that occur during accidents and maintenance and servicing
events. Finally, there are end-of-life emissions if, at the time of
vehicle scrappage, refrigerant is not fully recovered.
Because the process of refrigerant leakage has similar root causes
as those that cause fuel evaporative emissions from the fuel system,
some of the emission control technologies are similar (including hose
materials and connections). There are, however, some fundamental
differences between the systems that require a different approach, both
to controlling and to documenting that control. The most notable
difference is that A/C systems are completely closed systems and always
under significant pressure, whereas the fuel system is not. Fuel
systems are meant to be refilled as liquid fuel is consumed by the
engine, while the A/C system ideally should never require
``recharging'' of the contained refrigerant. Thus it is critical that
the A/C system leakages be kept to an absolute minimum. As a result,
these emissions are typically too low to accurately measure in most
current SHED chambers designed for fuel evaporative emissions
measurement, especially for A/C systems that are new or early in life.
A few commenters suggested that we allow manufacturers, as an
option, to use an industry-developed ``mini-shed'' test procedure (SAE
J2763--Test Procedure for Determining Refrigerant Emissions from Mobile
Air Conditioning Systems) to measure and report annual refrigerant
leakage.\206\ However, while EPA generally prefers performance testing,
for an individual vehicle A/C system or component, there is not a
strong inherent correlation between a performance test using SAE J2763
and the design-based approach we are adopting (based on SAE J2727, as
discussed below).\207\ Establishing such a correlation would require
testing of a fairly broad range of current-technology systems in order
to establish the effects of such factors as production variability and
assembly practices (which are included in J2727 scores, but not in
J2763 measurements). To EPA's knowledge, such a correlation study has
not been done. At the same time, as discussed below, there are
indications that much of the industry will eventually be moving toward
alternative refrigerants with very low GWPs. EPA believes such a
transition would diminish the value of any correlation

[[Page 25426]]

studies that might be done to confirm the appropriateness of the SAE
J2763 procedure as an option in this rule. For these reasons, EPA is
therefore not adopting such an optional direct measurement approach to
addressing refrigerant leakage at this time.
---------------------------------------------------------------------------

\206\ Honeywell and Volvo supported this view; most other
commenters did not.
\207\ However, there is a correlation in the fleet between J2763
measurements and J2727 scores.
---------------------------------------------------------------------------

Instead, as proposed, EPA is adopting a design-based method for
manufacturers to demonstrate improvements in their A/C systems and
components.\208\ Manufacturers implementing system designs expected to
result in reduced refrigerant leakage will be eligible for credits that
could then be used to meet their CO2 emission compliance
requirements (or otherwise banked or traded). The A/C Leakage Credit
provisions will generally assign larger credits to system designs that
would result in greater leakage reductions. In addition,
proportionately larger A/C Leakage Credits will be available to
manufacturers that substitute a refrigerant with lower GWP than the
current HFC-134a refrigerant.
---------------------------------------------------------------------------

\208\ See final regulations at 40 CFR 86.1866-12(b).
---------------------------------------------------------------------------

Our method for calculating A/C Leakage Credits is based closely on
an industry-consensus leakage scoring method, described below. This
leakage scoring method is correlated to experimentally-measured leakage
rates from a number of vehicles using the different available A/C
components. Under the approach, manufacturers will choose from a menu
of A/C equipment and components used in their vehicles in order to
establish leakage scores which will characterize their A/C system
leakage performance. Credits will be generated from leakage reduction
improvements that exceed average fleetwide leakage rates.
EPA believes that the design-based approach will result in
estimates of leakage emissions reductions that will be comparable to
those that will eventually result from performance-based testing. We
believe that this method appropriately approximates the real-world
leakage rates for the expected MY 2012-2016 A/C systems.
The cooperative industry and government Improved Mobile Air
Conditioning (IMAC) program \209\ has demonstrated that new-vehicle
leakage emissions can be reduced by 50%. This program has shown that
this level of improvement can be accomplished by reducing the number
and improving the quality of the components, fittings, seals, and hoses
of the A/C system. All of these technologies are already in commercial
use and exist on some of today's systems.
---------------------------------------------------------------------------

\209\ Team 1-Refrigerant Leakage Reduction: Final Report to
Sponsors, SAE, 2007.
---------------------------------------------------------------------------

As proposed, a manufacturer wishing to generate A/C Leakage Credits
will compare the components of its A/C system with a set of leakage-
reduction technologies and actions based closely on that developed
through IMAC and the Society of Automotive Engineers (as SAE Surface
Vehicle Standard J2727, August 2008 version). The J2727 approach was
developed from laboratory testing of a variety of A/C related
components, and EPA believes that the J2727 leakage scoring system
generally represents a reasonable correlation with average real-world
leakage in new vehicles. The EPA credit approach addresses the same A/C
components as does SAE J2727 and associates each component with the
same gram-per-year leakage rate as the SAE method, although, as
described below, EPA limits the credits allowed and also modifies it
for other factors such as alternative refrigerants.
A manufacturer choosing to generate A/C Leakage Credits will sum
the leakage values for an A/C system for a total A/C leakage score
according to the following formula. Because the primary GHG program
standards are expressed in terms of vehicle exhaust CO2
emissions as measured in grams per mile, the credits programs adopted
in this rule, including A/C related credits, must ultimately be
converted to a common metric for proper calculation of credits toward
compliance with the primary vehicle standards. This formula describes
the conversion of the grams-per-year leakage score to a grams-per-mile
CO2eq value, taking vehicle miles traveled (VMT) and the GWP
of the refrigerant into account:

A/C Leakage Credit = (MaxCredit) * [1-(LeakScore/AvgImpact) *
(GWPRefrigerant/1430)]

Where:

MaxCredit is 12.6 and 15.6 g/mi CO2eq for cars and
trucks, respectively. These values become 13.8 and 17.2 for cars and
trucks, respectively, if low-GWP refrigerants are used, since this
would generate additional credits from reducing emissions during
maintenance events, accidents, and at end-of-life.
LeakScore is the leakage score of the A/C system as measured
according to the EPA leakage method (based on the J2727 procedure,
as discussed above) in units of g/yr. The minimum score that EPA
considers feasible is fixed at 8.3 and 10.4 g/yr for cars and trucks
respectively (4.1 and 5.2 g/yr for systems using electric A/C
compressors) as discussed below.
Avg Impact is the average current A/C leakage emission rate, which
is 16.6 and 20.7 g/yr for cars and trucks, respectively.
GWPRefrigerant is the global warming potential (GWP) for direct
radiative forcing of the refrigerant. For purposes of this rule, the
GWP of HFC-134a is 1430, the GWP of HFC-152a is 124, the GWP of HFO-
1234yf is 4, and the GWP of CO2 as a refrigerant is 1.

The EPA Final RIA elaborates further on the development of each of
the values incorporated in the A/C Leakage Credit formula above, as
summarized here. First, as proposed, EPA estimates that leakage
emission rates for systems using the current refrigerant (HFC-134a)
could be feasibly reduced to rates no less than 50% of current rates--
or 8.3 and 10.4 g/yr for cars and trucks, respectively--based on the
conclusions of the IMAC study as well as consideration of refrigerant
emissions over the full life of the vehicle.
Also, some commenters noted that A/C compressors powered by
electric motors (e.g. as used today in several hybrid vehicle models)
were not included in the IMAC study and yet allow for leakage emission
rate reductions beyond EPA's estimates for systems with conventional
belt-driven compressors. EPA agrees with these comments, and we have
incorporated lower minimum emission rates into the formula above--4.1
and 5.2 g/yr for cars and trucks, respectively--in order to allow
additional leakage reduction credits for vehicles that use sealed
electric A/C compressors. The maximum available credits for these two
approaches are summarized in Table III.C.1-1 below.
AIAM commented that EPA should not set a lower limit on the leakage
score, even for non-electric compressors. EPA has determined not to do
so. First, although there do exist vehicles in the Minnesota data with
lower scores than our proposed (and now final) minimum scores, there
are very few car models that have scores less than 8.3, and these range
from 7.0 to about 8.0 and the difference are small compared to our
minimum score.\210\ More important, lowering the leakage limit would
necessarily increase credit opportunities for equipment design changes,
and EPA believes that these changes could discourage the
environmentally optimal result of using low GWP refrigerants.
Introduction of low GWP refrigerants could be discouraged because it
may be less costly to reduce leakage than to replace many of the A/C
system components. Moreover, due to the likelihood of in-use factors,
even a leakless (according to

[[Page 25427]]

J2727) R134a system will have some emissions due to manufacturing
variability, accidents, deterioration, maintenance, and end of life
emissions, a further reason to cap the amount of credits available
through equipment design. The only way to guarantee a near zero
emission system in-use is to use a low GWP refrigerant. The EPA has
therefore decided for the purposes of this final rule to not change the
minimum score for belt driven compressors due to the reason cited above
and to the otherwise overwhelming support for the program as proposed
from commenters.
---------------------------------------------------------------------------

\210\ The Minnesota refrigerant leakage data can be found at
http://www.pca.state.mn.us/climatechange/mobileair.html#leakdata.
---------------------------------------------------------------------------

In addition, as discussed above, EPA recognizes that substituting a
refrigerant with a significantly lower GWP will be a very effective way
to reduce the impact of all forms of refrigerant emissions, including
maintenance, accidents, and vehicle scrappage. To address future GHG
regulations in Europe and California, systems using alternative
refrigerants--including HFO1234yf, with a GWP of 4 and CO2
with a GWP of 1--are under serious development and have been
demonstrated in prototypes by A/C component suppliers. The European
Union has enacted regulations phasing in alternative refrigerants with
GWP less than 150 starting this year, and the State of California
proposed providing credits for alternative refrigerant use in its GHG
rule. Within the timeframe of MYs 2012-2016, EPA is not expecting
widespread use of low-GWP refrigerants. However, EPA believes that
these developments are promising, and, as proposed, has included in the
A/C Leakage Credit formula above a factor to account for the effective
GHG reductions that could be expected from refrigerant substitution.
The A/C Leakage Credits that will be available will be a function of
the GWP of the alternative refrigerant, with the largest credits being
available for refrigerants with GWPs at or approaching a value of 1.
For a hypothetical alternative refrigerant with a GWP of 1 (e.g.,
CO2 as a refrigerant), effectively eliminating leakage as a
GHG concern, our credit calculation method could result in maximum
credits equal to total average emissions, or credits of 13.8 and 17.2
g/mi CO2eq for cars and trucks, respectively, as
incorporated into the A/C Leakage Credit formula above as the
``MaxCredit'' term.
Table III.C.1-1 summarizes the maximum A/C leakage credits
available to a manufacturer, according to the formula above.

Table III.C.1-1--Maximum Leakage Credit Available to Manufacturers
------------------------------------------------------------------------
Car (g/mi) Truck (g/mi)
------------------------------------------------------------------------
R-134a refrigerant with belt- 6.3 7.8
driven compressor................
R-134a refrigerant with electric 9.5 11.7
motor-driven compressor..........
Lowest-GWP refrigerant (GWP=1).... 13.8 17.2
------------------------------------------------------------------------

It is possible that alternative refrigerants could, without
compensating action by the manufacturer, reduce the efficiency of the
A/C system (see related discussion of the A/C Efficiency Credit below.)
However, as noted at proposal and discussed further in the following
section, EPA believes that manufacturers will have substantial
incentives to design their systems to maintain the efficiency of the A/
C system. Therefore EPA is not accounting for any potential efficiency
degradation due to the use of alternative refrigerants.
Beyond the comments mentioned above, commenters generally supported
or were silent about EPA's refrigerant leakage methodology (as based on
SAE J2727), including the maximum leakage credits available, the
technologies eligible for credit and their associated leakage reduction
values, and the potential for alternative refrigerants. All comments
related to A/C credits are addressed in the Response to Comments
Document.
b. A/C Efficiency Credits
Manufacturers that make improvements in their A/C systems to
increase efficiency and thus reduce CO2 emissions due to A/C
system operation may be eligible for A/C Efficiency Credits. As with A/
C Leakage Credits, manufacturers could apply A/C Efficiency Credits
toward compliance with their overall CO2 standards (or
otherwise bank and trade the credits).
As mentioned above, EPA estimates that the CO2 emissions
due to A/C related loads on the engine account for approximately 3.9%
of total greenhouse gas emissions from passenger vehicles in the United
States. Usage of A/C systems is inherently higher in hotter and more
humid months and climates; however, vehicle owners may use their A/C
systems all year round in all parts of the nation. For example, people
commonly use A/C systems to cool and dehumidify the cabin air for
passenger comfort on hot humid days, but they also use the systems to
de-humidify cabin air to assist in defogging/de-icing the front
windshield and side glass in cooler weather conditions for improved
visibility. A more detailed discussion of seasonal and geographical A/C
usage rates can be found in the RIA.
Most of the additional load on the engine from A/C system operation
comes from the compressor, which pumps the refrigerant around the
system loop. Significant additional load on the engine may also come
from electric or hydraulic fans, which are used to move air across the
condenser, and from the electric blower, which is used to move air
across the evaporator and into the cabin. Manufacturers have several
currently-existing technology options for improving efficiency,
including more efficient compressors, fans, and motors, and system
controls that avoid over-chilling the air (and subsequently re-heating
it to provide the desired air temperature with an associated loss of
efficiency). For vehicles equipped with automatic climate-control
systems, real-time adjustment of several aspects of the overall system
(such as engaging the full capacity of the cooling system only when it
is needed, and maximizing the use of recirculated air) can result in
improved efficiency. Table III.C.1-2 below lists some of these
technologies and their respective efficiency improvements.
As discussed in the proposal, EPA is adopting a design-based
``menu'' approach for estimating efficiency improvements and, thus,
quantifying A/C Efficiency Credits.\211\ However, EPA's ultimate
preference is performance-based standards and credit mechanisms (i.e.,
using actual measurements) as typically providing a more accurate
measure of performance. However, EPA has concluded that a practical,
performance-based procedure for the purpose of accurately quantifying
A/C-related CO2 emission reductions, and thus efficiency
improvements for assigning credits, is not yet available. Still, EPA is
introducing a new specialized performance-based test for the more
limited purpose of demonstrating that

[[Page 25428]]

actual efficiency improvements are being achieved by the design
improvements for which a manufacturer is seeking A/C credits. As
discussed below, beginning in MY 2014, manufacturers wishing to
generate A/C Efficiency Credits will need to show improvement on the
new A/C Idle Test in order to then use the ``menu'' approach to
quantify the number of credits attributable to those improvements.
---------------------------------------------------------------------------

\211\ See final regulations at 40 CFR 86.1866-12(c).
---------------------------------------------------------------------------

In response to comments concerning the applicability and
effectiveness of technologies that were or were not included in our
analysis, we have made several changes to the design-based menu.\212\
First, we have separated the credit available for `recirculated air'
\213\ technologies into those with closed-loop control of the air
supply and those with open-loop control. By ``closed-loop'' control, we
mean a system that uses feedback from a sensor, or sensors, (e.g.,
humidity, glass fogging, CO2, etc.) to actively control the
interior air quality. For those systems that use ``open-loop'' control
of the air supply, we project that since this approach cannot precisely
adjust to varying ambient humidity or passenger respiration levels, the
relative effectiveness will be less than that for systems using closed-
loop control.
---------------------------------------------------------------------------

\212\ Commenters included the Alliance of Automobile
Manufacturers, Jaguar Land Rover, Denso, and the Motor and Equipment
Manufacturers Association, among others.
\213\ Recirculated air is defined as air present in the
passenger compartment of the vehicle (versus outside air) available
for the A/C system to cool or condition.
---------------------------------------------------------------------------

Second, many commenters indicated that the electronic expansion
valve, or EXV, should not be included in the menu of technologies, as
its effectiveness may not be as high as we projected. Commenters noted
that the SAE IMAC report stated efficiency improvements for an EXV used
in conjunction with a more efficient compressor, and not as a stand
alone technology and that no manufacturers are considering this
technology for their products within the timeframe of this rulemaking.
We believe other technologies (improved compressor controls for
example) can achieve the same benefit as an EXV, without the need for
this unique component, and therefore are not adopting it as an option
in the design menu of efficiency-improving A/C technologies.
Third, many commenters requested that an internal heat exchanger,
or IHX, be added to the design menu. EPA initially considered adding
this technology, but in our initial review of studies on this
component, we had understood that the value of the technology is
limited to systems using the alternative refrigerant HFO-1234yf. Some
manufacturers, however, commented that an IHX can also be used with
systems using the current refrigerant HFC-134a to improve efficiency,
and that they plan on implementing this technology as part their
strategy to improve A/C efficiency. Based on these comments, and
projections in a more recent SAE Technical Paper, we project that an
IHX in a conventional HFC-134a system can improve system efficiency by
20%, resulting in a credit of 1.1 g/mi.\214\ Further discussion of IHX
technology can be found in the RIA.
---------------------------------------------------------------------------

\214\ Mathur, Gursaran D., ``Experimental Investigation with
Cross Fluted Double-Pipe Suction Line Heat Exchanger to Enhance A/C
System Performance,'' SAE 2009-01-0970, 2009.
---------------------------------------------------------------------------

Fourth, we have modified the definition of `improved evaporators
and condensers' to recognize that improved versions of these heat
exchangers may be used separately or in conjunction with one another,
and that an engineering analysis must indicate a COP improvement of 10%
or better when using either or both components (and not a 10% COP
improvement for each component). Furthermore, we have modified the
regulation text to clarify what is considered to be the `baseline'
components for this analysis. We consider the baseline component to be
the version which a manufacturer most recently had in production on the
same vehicle or a vehicle in a similar EPA vehicle classification. The
dimensional characteristics (e.g. tube configuration/thickness/spacing,
and fin density) of the baseline components are then compared to the
new components, and an engineering analysis is required to demonstrate
the COP improvement.
For model years 2012 and 2013, a manufacturer wishing to generate
A/C Efficiency Credits for a group of its vehicles with similar A/C
systems will compare several of its vehicle A/C-related components and
systems with a list of efficiency-related technology improvements (see
Table III.C.1-2 below). Based on the technologies the manufacturer
chooses, an A/C Efficiency Credit value will be established. This
design-based approach will recognize the relationships and synergies
among efficiency-related technologies. Manufacturers could receive
credits based on the technologies they chose to incorporate in their A/
C systems and the associated credit value for each technology. The
total A/C Efficiency Credit will be the total of these values, up to a
maximum allowable credit of 5.7 g/mi CO2eq. This will be the
maximum improvement from current average efficiencies for A/C systems
(see the RIA for a full discussion of our derivation of the reductions
and credit values for individual technologies and for the maximum total
credit available). Although the total of the individual technology
credit values may exceed 5.7 g/mi CO2eq, synergies among the
technologies mean that the values are not additive. A/C Efficiency
Credits as adopted may not exceed 5.7 g/mi CO2eq.

Table III.C.1-2--Efficiency-Improving A/C Technologies and Credits
------------------------------------------------------------------------
Estimated
reduction in A/C A/C efficiency
Technology description CO2 emissions credit (g/mi
(%) CO2)
------------------------------------------------------------------------
Reduced reheat, with externally- 30 1.7
controlled, variable-displacement
compressor.........................
Reduced reheat, with externally- 20 1.1
controlled, fixed-displacement or
pneumatic variable-displacement
compressor.........................
Default to recirculated air with 30 1.7
closed-loop control of the air
supply (sensor feedback to control
interior air quality) whenever the
ambient temperature is 75 [deg]F or
higher (although deviations from
this temperature are allowed if
accompanied by an engineering
analysis)..........................
Default to recirculated air with 20 1.1
open-loop control air supply (no
sensor feedback) whenever the
ambient temperature 75 [deg]F or
higher lower temperatures are
allowed............................
Blower motor controls which limit 15 0.9
wasted electrical energy (e.g.,
pulse width modulated power
controller)........................
Internal heat exchanger............. 20 1.1
Improved condensers and/or 20 1.1
evaporators (with system analysis
on the component(s) indicating a
COP improvement greater than 10%,
when compared to previous industry
standard designs)..................

[[Page 25429]]

Oil separator (with engineering 10 0.6
analysis demonstrating
effectiveness relative to the
baseline design)...................
------------------------------------------------------------------------

The proposal requested comment on adjusting the efficiency credit
for alternative refrigerants. Although a few commenters noted that the
efficiency of an HFO1234yf system may differ from a current HFC-134a
system,\215\ we believe that this difference does not take into account
any efficiency improvements that may be recovered or gained when the
overall system is specifically designed with consideration of the new
refrigerant properties (as compared to only substituting the new
refrigerant). EPA is therefore not adjusting the credits based on
efficiency differences for this rule.
---------------------------------------------------------------------------

\215\ Ford noted that ``the physical properties of the
alternative refrigerant R1234yf could result in a reduction of
efficiency by 5 to 10 percent compared to R134a in use today with a
similar refrigerant system and controls technology.''
---------------------------------------------------------------------------

As noted above, for model years 2014 and later, manufacturers
seeking to generate design-based A/C Efficiency Credits will also need
to use a specific new EPA performance test to confirm that the design
changes are resulting in improvements in A/C system efficiency as
integrated into the vehicle. As proposed, beginning in MY 2014
manufacturers will need to perform an A/C CO2 Idle Test for
each A/C system (family) for which it desires to generate Efficiency
Credits. Manufacturers will need to demonstrate an improvement over
current average A/C CO2 levels (21.3 g/minute on the Idle
Test) to qualify for the menu approach credits. Upon qualifying on the
Idle Test, the manufacturer will be eligible to use the menu approach
above to quantify the potential credits it could generate. To earn the
full amount of credits available in the menu approach (limited to the
maximum), the test must demonstrate a 30% or greater improvement in
CO2 levels over the current average.
For A/C systems that achieve an improvement between 0-and-30% (or a
result between 21.3 and 14.9 g/minute result on the A/C CO2
Idle Test), a credit can still be earned, but a multiplicative credit
adjustment factor will be applied to the eligible credits. As shown in
Figure III.C.1-1 this factor will be scaled from 1.0 to 0, with
vehicles demonstrating a 30% or better improvement (14.9 g/min or
lower) receiving 100% of the eligible credit (adj. factor = 1.0), and
vehicles demonstrating a 0% improvement--21.3 g/min or higher result--
receiving no credit (adj. factor = 0). We adopted this adjustment
factor in response to commenters who were concerned that a vehicle
which incorporated many efficiency-improving technologies may not
achieve the full 30% improvement, and as a result would receive no
credit (thus discouraging them from using any of the technologies).
Because there is environmental benefit (reduced CO2) from
the use of even some of these efficiency-improving technologies, EPA
believes it is appropriate to scale the A/C efficiency credits to
account for these partial improvements.
BILLING CODE 6560-50-P

[[Page 25430]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.016

BILLING CODE 6560-50-C

[[Page 25431]]

EPA is adopting the A/C CO2 Idle Test procedure as
proposed in most respects. This laboratory idle test is performed while
the vehicle is at idle, similar to the idle carbon monoxide (CO) test
that was once a part of EPA vehicle certification. The test determines
the additional CO2 generated at idle when the A/C system is
operated. The A/C CO2 Idle Test will be run with and without
the A/C system cooling the interior cabin while the vehicle's engine is
operating at idle and with the system under complete control of the
engine and climate control system. The test includes tighter
restrictions on test cell temperatures and humidity levels than apply
for the basic FTP test procedure in order to more closely control the
loads from operation of the A/C system. EPA is also adopting additional
refinements to the required in-vehicle blower fan settings for manually
controlled systems to more closely represent ``real world'' usage
patterns.
Many commenters questioned the ability of this test to measure the
improved efficiency of certain A/C technologies, and stated that the
test was not representative of real-world driving conditions. However,
although EPA acknowledges that this test directly simulates a
relatively limited range of technologies and conditions, we determined
that it is sufficiently robust for the purpose of demonstrating that
the system design changes are indeed implemented properly and are
resulting in improved efficiency of a vehicle's A/C system, at idle as
well as under a range of operating conditions. Further details of the
A/C Idle Test can be found in the RIA and the regulations, as well as
in the Response to Comments Document.
The design of the A/C CO2 Idle Test represents a
balancing of the need for performance tests whenever possible to ensure
the most accurate quantification of efficiency improvements, with
practical concerns for testing burden and facility requirements. EPA
believes that the Idle Test adds to the robust quantification of A/C
credits that will result in real-world efficiency improvements and
reductions in A/C-related CO2 emissions. The Idle Test will
not be required in order to generate A/C Efficiency Credits until MY
2014 to allow sufficient time for manufacturers to make the necessary
facilities improvements and to gain experience with the test.
EPA also considered and invited comment on a more comprehensive
testing approach to quantifying A/C CO2 emissions that could
be somewhat more technically robust, but would require more test time
and test facility improvements for many manufacturers. EPA invited
comment on using an adapted version of the SCO3, an existing test
procedure that is part of the Supplemental Federal Test Procedure. EPA
discussed and invited comment on the various benefits and concerns
associated with using an adapted SCO3 test. There were many comments
opposed to this proposal, and very few supporters. Most of the comments
opposing this approach echoed the concerns made by in the NPRM. These
included excessive testing burden, limited test facilities and the cost
of adding new ones, and the concern that the SC03 test may not be
sufficiently representative of in use A/C usage. Some commenters
supported a derivative of the SCO3 test or multiple runs of other urban
cycles (such as the LA-4) for quantifying A/C system efficiency. While
EPA considers a test cycle that covers a broader range of vehicle speed
and climatic conditions to be ideal, developing such a representative
A/C test would involve the work of many stakeholders, and would require
a significant amount of time, exceeding the scope of this rule. EPA
expects to continue working with industry, the California Air Resources
Board, and other stakeholders to move toward increasingly robust
performance tests and methods for determining the efficiency of mobile
A/C systems and the related impact on vehicle CO2 emissions,
including a potential adapted SC03 test.
c. Interaction With Title VI Refrigerant Regulations
Title VI of the Clean Air Act deals with the protection of
stratospheric ozone. Section 608 establishes a comprehensive program to
limit emissions of certain ozone-depleting substances (ODS). The rules
promulgated under section 608 regulate the use and disposal of such
substances during the service, repair or disposal of appliances and
industrial process refrigeration. In addition, section 608 and the
regulations promulgated under it, prohibit knowingly venting or
releasing ODS during the course of maintaining, servicing, repairing or
disposing of an appliance or industrial process refrigeration
equipment. Section 609 governs the servicing of motor vehicle A/C
systems. The regulations promulgated under section 609 (40 CFR part 82,
subpart B) establish standards and requirements regarding the servicing
of A/C systems. These regulations include establishing standards for
equipment that recovers and recycles (or, for refrigerant blends, only
recovers) refrigerant from A/C systems; requiring technician training
and certification by an EPA-approved organization; establishing
recordkeeping requirements; imposing sales restrictions; and
prohibiting the venting of refrigerants. Section 612 requires EPA to
review substitutes for class I and class II ozone depleting substances
and to consider whether such substitutes will cause an adverse effect
to human health or the environment as compared with other substitutes
that are currently or potentially available. EPA promulgated
regulations for this program in 1992 and those regulations are located
at 40 CFR part 82, subpart G. When reviewing substitutes, in addition
to finding them acceptable or unacceptable, EPA may also find them
acceptable so long as the user meets certain use conditions. For
example, all motor vehicle air conditioning systems must have unique
fittings and a uniquely colored label for the refrigerant being used in
the system.
On September 14, 2006, EPA proposed to approve R-744
(CO2) for use in motor vehicle A/C systems (71 FR 55140) and
on October 19, 2009, EPA proposed to approve the low-GWP refrigerant
HFO-1234yf for these systems (74 FR 53445), both subject to certain
requirements. Final action on both of these proposals is expected later
this year. EPA previously issued a final rule allowing the use of HFC-
152a as a refrigerant in motor vehicle A/C systems subject to certain
requirements (June 12, 2008; 73 FR 33304). As discussed above,
manufacturers transitioning to any of the approved refrigerants would
be eligible for A/C Leakage Credits, the value of which would depend on
the GWP of their refrigerant and the degree of leakage reduction of
their systems.
EPA views this rule as complementing these Title VI programs, and
not conflicting with them. To the extent that manufacturers choose to
reduce refrigerant leakage in order to earn A/C Leakage Credits, this
will dovetail with the Title VI section 609 standards which apply to
maintenance events, and to end-of-vehicle life disposal. In fact, as
noted, a benefit of the A/C credit provisions is that there should be
fewer and less impactive maintenance events for MVACs, since there will
be less leakage. In addition, the credit provisions will not conflict
(or overlap) with the Title VI section 609 standards. EPA also believes
the menu of leak control technologies described in this rule will
complement the section 612 requirements, because these control
technologies will help ensure that HFC-134a (or other refrigerants)
will be used in a manner that further minimizes potential adverse

[[Page 25432]]

effects on human health and the environment.
2. Flexible Fuel and Alternative Fuel Vehicle Credits
EPA is finalizing its proposal to allow flexible-fuel vehicles
(FFVs) and alternative fuel vehicles to generate credits for purposes
of the GHG rule starting in the 2012 model year. FFVs are vehicles that
can run on both an alternative fuel and a conventional fuel. Most FFVs
are E85 vehicles, which can run on a mixture of up to 85 percent
ethanol and gasoline. Dedicated alternative fuel vehicles are vehicles
that run exclusively on an alternative fuel (e.g., compressed natural
gas). These credits are designed to complement the treatment of FFVs
under CAFE, consistent with the emission reduction objectives of the
CAA. As explained at proposal, EPCA includes an incentive under the
CAFE program for production of dual-fueled vehicles or FFVs, and
dedicated alternative fuel vehicles.\216\ For FFVs and dual-fueled
vehicles, the EPCA/EISA credits have three elements: (1) The assumption
that the vehicle is operated 50% of the time on the conventional fuel
and 50% of the time on the alternative fuel, (2) that 1 gallon of
alternative fuel is treated as 0.15 gallon of fuel, essentially
increasing the fuel economy of a vehicle on alternative fuel by a
factor of 6.67, and (3) a ``cap'' provision that limits the maximum
fuel economy increase that can be applied to a manufacturer's overall
CAFE compliance value for all CAFE compliance categories (i.e.,
domestic passenger cars, import passenger cars, and light trucks) to
1.2 mpg through 2014 and 1.0 mpg in 2015. EPCA's provisions were
amended by the EISA to extend the period of availability of the FFV
credits, but to begin phasing them out by annually reducing the amount
of FFV credits that can be used in demonstrating compliance with the
CAFE standards.\217\ EPCA does not premise the availability of the FFV
credits on actual use of alternative fuel. Under EPCA, after MY 2019 no
FFV credits will be available for CAFE compliance.\218\ Under EPCA, for
dedicated alternative fuel vehicles, there are no limits or phase-out.
As proposed, FFV and Alternative Fuel Vehicle Credits will be
calculated as a part of the calculation of a manufacturer's overall
fleet average fuel economy and fleet average carbon-related exhaust
emissions (Sec. 600.510-12).
---------------------------------------------------------------------------

\216\ 49 U.S.C. 32905.
\217\ See 49 U.S.C. 32906. The mechanism by which EPCA provides
an incentive for production of FFVs is by specifying that their fuel
economy is determined using a special calculation procedure that
results in those vehicles being assigned a higher fuel economy level
than would otherwise occur. 49 U.S.C. 32905(b). This is typically
referred to as an FFV credit.
\218\ 49 U.S.C. 32906.
---------------------------------------------------------------------------

Manufacturers supported the inclusion of FFV credits in the
program. Chrysler noted that the credits encourage manufacturers to
continue production of vehicles capable of running on alternative fuels
as the production and distribution systems of such fuels are developed.
Chrysler believes the lower carbon intensity of such fuels is an
opportunity for further greenhouse gas reductions and increased energy
independence, and the continuance of such incentives recognizes the
important potential of this technology to reduce GHGs. Toyota noted
that because actions taken by manufacturers to comply with EPA's
regulation will, to a large extent, be the same as those taken to
comply with NHTSA's CAFE regulation, it is appropriate for EPA to
consider flexibilities contained in the CAFE program that clearly
impact product plans and technology deployment plans already in place
or nearly in place. Toyota believes that adopting the FFV credit for a
transitional period of time appears to recognize this reality, while
providing a pathway to eventually phase-out the flexibility.
As proposed, electric vehicles (EVs) or plug-in hybrid electric
vehicles (PHEVs) are not eligible to generate this type of credit.
These vehicles are covered by the advanced technology vehicle
incentives provisions described in Section III.C.3, so including them
here would lead to a double counting of credits.
a. Model Year 2012-2015 Credits
i. FFVs
For the GHG program, EPA is allowing FFV credits corresponding to
the amounts allowed by the amended EPCA but only during the period from
MYs 2012 to 2015. (As discussed below in Section III.E., EPA is not
allowing CAFE-based FFV credits to be generated as part of the early
credits program.) As noted at proposal, several manufacturers have
already taken the availability of FFV credits into account in their
near-term future planning for CAFE and this reliance indicates that
these credits need to be considered in assessing necessary lead time
for the CO2 standards. Manufacturers commented that the
credits are necessary in allowing them to transition to the new
standards. EPA thus believes that allowing these credits, in the near
term, would help provide adequate lead time for manufacturers to
implement the new multi-year standards, but that for the longer term
there is adequate lead time without the use of such credits. This will
also tend to harmonize the GHG and the CAFE program during these
interim years. As discussed below, EPA is requiring for MY 2016 and
later that manufacturers will need to reliably estimate the extent to
which the alternative fuel is actually being used by vehicles in order
to count the alternative fuel use in the vehicle's CO2
emissions level determination. Beginning in MY 2016, the FFV credits as
described above for MY 2012-2015 will no longer be available for EPA's
GHG program. Rather, GHG compliance values will be based on actual
emissions performance of the FFV on conventional and alternative fuels,
weighted by the actual use of these fuels in the FFVs.
As with the CAFE program, EPA will base MY 2012-2015 credits on the
assumption that the vehicles would operate 50% of the time on the
alternative fuel and 50% of the time on conventional fuel, resulting in
CO2 emissions that are based on an arithmetic average of
alternative fuel and conventional fuel CO2 emissions.\219\
In addition, the measured CO2 emissions on the alternative
fuel will be multiplied by a 0.15 volumetric conversion factor which is
included in the CAFE calculation as provided by EPCA. Through this
mechanism a gallon of alternative fuel is deemed to contain 0.15
gallons of fuel. For example, for a flexible-fuel vehicle that emitted
330 g/mi CO2 operating on E85 and 350 g/mi CO2
operating on gasoline, the resulting CO2 level to be used in
the manufacturer's fleet average calculation would be:
---------------------------------------------------------------------------

\219\ 49 U.S.C. 32905(b).
[GRAPHIC] [TIFF OMITTED] TR07MY10.017

EPA understands that by using the CAFE approach--including the 0.15
factor--the CO2 emissions value for the vehicle is
calculated to be significantly lower than it actually would be
otherwise, even if the vehicle were assumed to operate on the
alternative fuel at all times. This represents a ``credit'' being
provided to FFVs.
EPA notes also that the above equation and example are based on an
FFV that is an E85 vehicle. EPCA, as amended by EISA, also establishes
the use of this approach, including the 0.15 factor, for all
alternative fuels, not just

[[Page 25433]]

E85.\220\ The 0.15 factor is used for B-20 (20 percent biofuel and 80
percent diesel) FFVs. EPCA also establishes this approach, including
the 0.15 factor, for gaseous-fueled dual-fueled vehicles, such as a
vehicle able to operate on gasoline and CNG.\221\ (For natural gas
dual-fueled vehicles, EPCA establishes a factor of 0.823 gallons of
fuel for every 100 cubic feet a natural gas used to calculate a gallons
equivalent.\222\) The EISA's use of the 0.15 factor in this way
provides a similar regulatory treatment across the various types of
alternative fuel vehicles. EPA also will use the 0.15 factor for all
FFVs in order not to disrupt manufacturers' near-term compliance
planning and assure sufficient lead time. EPA, in any case, expects the
vast majority of FFVs to be E85 vehicles, as is the case today.
---------------------------------------------------------------------------

\220\ 49 U.S.C. 32905(c).
\221\ 49 U.S.C. 32905(d).
\222\ 49 U.S.C. 32905(c).
---------------------------------------------------------------------------

The FFV credit limits for CAFE are 1.2 mpg for model years 2012-
2014 and 1.0 mpg for model year 2015.\223\ In CO2 terms,
these CAFE limits translate to declining CO2 credit limits
over the four model years, as the CAFE standards increase in
stringency. As the CAFE standard increases numerically, the limit
becomes a smaller fraction of the standard. EPA proposed, but is not
adopting, credit limits based on the overall industry average
CO2 standards for cars and trucks. EPA also requested
comments on basing the calculated CO2 credit limits on the
individual manufacturer fleet-average standards calculated from the
footprint curves. EPA received comment from one manufacturer supporting
this approach. EPA also received comments from another manufacturer
recommending that the credit limits for an individual manufacturer be
based instead on that manufacturer's fleet average performance. The
commenter noted that this approach is in line with how CAFE FFV credit
limits are applied. This is due to the fact that the GHG-equivalent of
the CAFE 1.2 mpg cap will vary due to the non-linear relationship
between fuel economy and GHGs/fuel consumption. EPA agrees with this
approach since it best harmonizes how credit limits are determined in
CAFE. EPA intended and continues to believe it is appropriate to
provide essentially the same FFV credits under both programs for MYs
2012-2015. Therefore, EPA is finalizing FFV credits limits for MY 2012-
2015 based on a manufacturer's fleet-average performance. For example,
if a manufacturer's 2012 car fleet average emissions performance was
260 g/mile (34.2 mpg), the credit limit in CO2 terms would
be 9.5 g/mile (34.2 mpg - 1.2 mpg = 33.0 mpg = 269.5 g/mile) and if it
were 270 g/mile the limit would be 10.2 g/mile.
---------------------------------------------------------------------------

\223\ 49 U.S.C. 32906(a).
---------------------------------------------------------------------------

ii. Dedicated Alternative Fuel Vehicles
As proposed, EPA will calculate CO2 emissions from
dedicated alternative fuel vehicles for MY 2012-2015 by measuring the
CO2 emissions over the test procedure and multiplying the
results by the 0.15 conversion factor described above. For example, for
a dedicated alternative fuel vehicle that would achieve 330 g/mi
CO2 while operating on alcohol (ethanol or methanol), the
effective CO2 emissions of the vehicle for use in
determining the vehicle's CO2 emissions would be calculated
as follows:

CO2 = 330 x 0.15 = 49.5 g/mi
b. Model Years 2016 and Later
i. FFVs
EPA is treating FFV credits the same as under EPCA for model years
2012-2015, but is applying a different approach starting with model
year 2016. EPA recognizes that under EPCA automatic FFV credits are
entirely phased out of the CAFE program by MY 2020, and apply in the
prior model years with certain limitations, but without a requirement
that the manufacturers demonstrate actual use of the alternative fuel.
Unlike EPCA, CAA section 202(a) does not mandate that EPA treat FFVs in
a specific way. Instead EPA is required to exercise its own judgment
and determine an appropriate approach that best promotes the goals of
this CAA section. Under these circumstances, EPA will treat FFVs for
model years 2012-2015 the same as under EPCA, as part of providing
sufficient lead time given manufacturers' compliance strategies which
rely on the existence of these EPCA statutory credits, as explained
above.
Starting with model year 2016, as proposed, EPA will no longer
allow manufacturers to base FFV emissions on the use of the 0.15 factor
credit described above, and on the use of an assumed 50% usage of
alternative fuel. Instead, EPA believes the appropriate approach is to
ensure that FFV emissions are based on demonstrated emissions
performance. This will promote the environmental goals of the final
program. EPA received several comments in support of EPA's proposal to
use this approach instead of the EPCA approach for MY 2016 and later.
Under the EPA program in MY 2016 and later, manufacturers will be
allowed to base an FFV's emissions compliance value in part on the
vehicle test values run on the alternative fuel, for that portion of
its fleet for which the manufacturer demonstrates utilized the
alternative fuel in the field. In other words, the default is to assume
FFVs operate on 100% gasoline, and the emissions value for the FFV
vehicle will be based on the vehicle's tested value on gasoline.
However, if a manufacturer can demonstrate that a portion of its FFVs
are using an alternative fuel in use, then the FFV emissions compliance
value can be calculated based on the vehicle's tested value using the
alternative fuel, prorated based on the percentage of the fleet using
the alternative fuel in the field. An example calculation is described
below. EPA believes this approach will provide an actual incentive to
ensure that such fuels are used. The incentive arises since actual use
of the flexible fuel typically results in lower tailpipe GHG emissions
than use of gasoline and hence improves the vehicles' performance,
making it more likely that its performance will improve a
manufacturers' average fleetwide performance. Based on existing
certification data, E85 FFV CO2 emissions are typically
about 5 percent lower on E85 than CO2 emissions on 100
percent gasoline. Moreover, currently there is little incentive to
optimize CO2 performance for vehicles when running on E85.
EPA believes the above approach would provide such an incentive to
manufacturers and that E85 vehicles could be optimized through engine
redesign and calibration to provide additional CO2
reductions.
Under the EPCA credit provisions, there is an incentive to produce
FFVs but no actual incentive to ensure that the alternative fuels are
used, or that actual vehicle fuel economy improves. GHG and energy
security benefits are only achieved if the alternative fuel is actually
used and (for GHGs) that performance improves, and EPA's approach for
MY 2016 and beyond will now provide such an incentive. This approach
will promote greater use of alternative fuels, as compared to a
situation where there is a credit but no usage requirement. This is
also consistent with the agency's overall commitment to the expanded
use of renewable fuels. Therefore, EPA is basing the FFV program for
MYs 2016 and thereafter on real-world reductions: i.e., actual vehicle
CO2 emissions levels based on actual use of the two fuels,
without the 0.15 conversion factor specified under EISA.

[[Page 25434]]

For 2016 and later model years, EPA will therefore treat FFVs
similarly to conventional fueled vehicles in that FFV emissions would
be based on actual CO2 results from emission testing on the
fuels on which it operates. In calculating the emissions performance of
an FFV, manufacturers may base FFV emissions on vehicle testing based
on the alternative fuel emissions, if they can demonstrate that the
alternative fuel is actually being used in the vehicles. Performance
will otherwise be calculated assuming use only of conventional fuel.
The manufacturer must establish the ratio of operation that is on the
alternative fuel compared to the conventional fuel. The ratio will be
used to weight the CO2 emissions performance over the 2-
cycle test on the two fuels. The 0.15 conversion factor will no longer
be included in the CO2 emissions calculation. For example,
for a flexible-fuel vehicle that emitted 300 g/mi CO2
operating on E85 ten percent of the time and 350 g/mi CO2
operating on gasoline ninety percent of the time, the CO2
emissions for the vehicles to be used in the manufacturer's fleet
average would be calculated as follows:

CO2 = (300 x 0.10) + (350 x 0.90) = 345 g/mi

The most complex part of this approach is to establish what data
are needed for a manufacturer to accurately demonstrate use of the
alternative fuel, where the manufacturer intends for its performance to
be calculated based on some use of alternative fuels. One option EPA is
finalizing is establishing a rebuttable presumption using a national
average approach based on national E85 fuel use. Manufacturers could
use this value along with their vehicle emissions results demonstrating
lower emissions on E85 to determine the emissions compliance values for
FFVs sold by manufacturers under this program. For example, national
E85 volumes and national FFV sales may be used to prorate E85 use by
manufacturer sales volumes and FFVs already in-use. Upon a
manufacturer's written request, EPA will conduct an analysis of vehicle
miles travelled (VMT) by year for all FFVs using its emissions
inventory MOVES model. Using the VMT ratios and the overall E85 sales,
E85 usage will be assigned to each vehicle. This method accounts for
the VMT of new FFVs and FFVs already in the existing fleet using VMT
data in the model. The model will then be used to determine the ratio
of E85 and gasoline for new vehicles being sold. Fluctuations in E85
sales and FFV sales will be taken into account to adjust the
manufacturers' E85 actual use estimates annually. EPA plans to make
this assigned fuel usage factor available through guidance prior to the
start of MY 2016 and adjust it annually as necessary. EPA believes this
is a reasonable way to apportion E85 use across the fleet.
If manufacturers decide not to use EPA's assigned fuel usage based
on the national average analysis, they have a second option of
presenting their own data for consideration as the basis for evaluating
fuel usage. Manufacturers have suggested demonstrations using vehicle
on-board data gathering through the use of on-board sensors and
computers. California's program allows FFV credits based on FFV use and
envisioned manufacturers collecting fuel use data from vehicles in
fleets with on-site refueling. Manufacturers must present a statistical
analysis of alternative fuel usage data collected on actual vehicle
operation. EPA is not attempting to specify how the data is collected
or the amount of data needed. However, the analysis must be based on
sound statistical methodology. Uncertainty in the analysis must be
accounted for in a way that provides reasonable certainty that the
program does not result in loss of emissions reductions.
EPA received comments that the 2016 and later FFV emissions
performance methodology should be based on the life cycle emissions
(i.e., including the upstream GHG emissions associated with fuel
feedstocks, production, and transportation) associated with the use of
the alternative fuel. Commenters are concerned that the use of ethanol
will not result in lower GHGs on a lifecycle basis. After considering
these comments, EPA is not including lifecycle emissions in the
calculation of vehicle credits. EPA continues to believe that it is
appropriate to base credits for MY 2012-2015 on the EPCA/CAFE credits
and to base compliance values for MY 2016 on the demonstrated tailpipe
emissions performance on gasoline and E85, and is finalizing this
approach as proposed. EPA recently finalized its RFS2 rulemaking which
addresses lifecycle emissions from ethanol and the upstream GHG
benefits of E85 use are already captured by this program.\224\
---------------------------------------------------------------------------

\224\ 75 FR 14670 (March 26, 2010).
---------------------------------------------------------------------------

ii. Dedicated Alternative Fuel Vehicles
As proposed, for model years 2016 and later dedicated alternative
fuel vehicles, CO2 will be measured over the 2-cycle test in
order to be included in a manufacturer's fleet average CO2
calculations. As noted above, this is different than CAFE methodology
which provides a methodology for calculating a petroleum-based mpg
equivalent for alternative fuel vehicles so they can be included in
CAFE. However, because CO2 can be measured directly from
alternative fuel vehicles over the test procedure, EPA believes this is
the simplest and best approach since it is consistent with all other
vehicle testing under the CO2 program. EPA did not receive
comments on this approach.
3. Advanced Technology Vehicle Incentives for Electric Vehicles, Plug-
in Hybrids, and Fuel Cell Vehicles
EPA is finalizing provisions that provide a temporary regulatory
incentive for the commercialization of certain advanced vehicle power
trains--electric vehicles (EVs), plug-in hybrid electric vehicles
(PHEVs), and fuel cell vehicles (FCVs)--for model year 2012-2016 light-
duty and medium-duty passenger vehicles.\225\ The purpose of these
provisions is to provide a temporary incentive to promote technologies
which have the potential to produce very large GHG reductions in the
future, but which face major challenges such as vehicle cost, consumer
acceptance, and the development of low-GHG fuel production
infrastructure. The tailpipe GHG emissions from EVs, PHEVs operated on
grid electricity, and hydrogen-fueled FCVs are zero, and traditionally
the emissions of the vehicle itself are all that EPA takes into account
for purposes of compliance with standards set under section 202(a).
Focusing on vehicle tailpipe emissions has not raised any issues for
criteria pollutants, as upstream emissions associated with production
and distribution of the fuel are addressed by comprehensive regulatory
programs focused on the upstream sources of those emissions.\226\ At
this time, however, there is no such comprehensive program addressing
upstream emissions of GHGs, and the upstream GHG emissions associated
with production and distribution of electricity are higher than the
corresponding upstream GHG emissions of gasoline or other petroleum
based fuels. In the future, if there were a program to comprehensively
control upstream GHG emissions, then the zero tailpipe levels from
these vehicles have the potential to produce very large GHG reductions,
and to transform the

[[Page 25435]]

transportation sector's contribution to nationwide GHG emissions.
---------------------------------------------------------------------------

\225\ See final regulations at 40 CFR 86.1866-12(a).
\226\ In this section, ``upstream'' means all fuel-related GHG
emissions prior to the fuel being introduced to the vehicle.
---------------------------------------------------------------------------

This temporary incentive program applies only for the model years
2012-2016 covered by this final rule. EPA will reassess the issue of
how to address EVs, PHEVs, and FCVs in rulemakings for model years 2017
and beyond, based on the status of advanced technology vehicle
commercialization, the status of upstream GHG emissions control
programs, and other relevant factors.
In the Joint Notice of Intent, EPA stated that ``EPA is currently
considering proposing additional credit opportunities to encourage the
commercialization of advanced GHG/fuel economy control technology such
as electric vehicles and plug-in hybrid electric vehicles. These `super
credits' could take the form of a multiplier that would be applied to
the number of vehicles sold such that they would count as more than one
vehicle in the manufacturer's fleet average.'' \227\ Following through,
EPA proposed two mechanisms by which these vehicles would earn credits:
(1) A zero grams/mile compliance value for EVs, FCVs, and for PHEVs
when operated on grid electricity, and (2) a vehicle multiplier in the
range of 1.2 to 2.0.\228\
---------------------------------------------------------------------------

\227\ Notice of Upcoming Joint Rulemaking to Establish Vehicle
GHG Emissions and CAFE Standards, 74 FR 24007, 24011 (May 22, 2009).
\228\ 74 FR 49533-34.
---------------------------------------------------------------------------

The zero grams/mile compliance value for EVs (and for PHEVs when
operated on grid electricity, as well as for FCVs which involve similar
upstream GHG issues with respect to hydrogen production) is an
incentive that operates like a credit because, while it accurately
accounts for tailpipe GHG emissions, it does not reflect the increase
in upstream GHG emissions associated with the electricity used by EVs
compared to the upstream GHG emissions associated with the gasoline or
diesel fuel used by conventional vehicles.\229\ For example, based on
GHG emissions from today's national average electricity generation
(including GHG emissions associated with feedstock extraction,
processing, and transportation) and other key assumptions related to
vehicle electricity consumption, vehicle charging losses, and grid
transmission losses, a midsize EV might have an upstream GHG emissions
of about 180 grams/mile, compared to the upstream GHG emissions of a
typical midsize gasoline car of about 60 grams/mile. Thus, the EV would
cause a net upstream GHG emissions increase of about 120 grams/mile (in
general, the net upstream GHG increase would be less for a smaller EV
and more for a larger EV). The zero grams/mile compliance value
provides an incentive because it is less than the 120 grams/mile value
that would fully account for the net increase in GHG emissions,
counting upstream emissions.\230\ The net upstream GHG impact could
change over time, of course, based on changes in electricity generation
or gasoline production.
---------------------------------------------------------------------------

\229\ See 74 FR 49533 (``EPA recognizes that for each EV that is
sold, in reality the total emissions off-set relative to the typical
gasoline or diesel powered vehicle is not zero, as there is a
corresponding increase in upstream CO2 emissions due to
an increase in the requirements for electric utility generation'').
\230\ This 120 grams/mile value for a midsize EV is
approximately similar to the compliance value for today's most
efficient conventional hybrid vehicle, so the EV would not be
significantly more ``GHG-positive'' than the most efficient
conventional hybrid counterpart under a full accounting approach. It
should be noted that these emission levels would still be well below
the footprint targets for the vehicles in question.
---------------------------------------------------------------------------

The proposed vehicle multiplier incentive would also have operated
like a credit as it would have allowed an EV, PHEV, or FCV to count as
more than one vehicle in the manufacturer's fleet average. For example,
combining a multiplier of 2.0 with a zero grams/mile compliance value
for an EV would allow that EV to be counted as two vehicles, each with
a zero grams/mile compliance value, in the manufacturer's fleet average
calculations. In effect, a multiplier of 2.0 would double the overall
credit associated with an EV, PHEV, or FCV.
EPA explained in the proposal that the potential for large future
emissions benefits from these technologies provides a strong reason for
providing incentives at this time to promote their commercialization in
the 2012-2016 model years. At the same time, EPA acknowledged that the
zero grams/mile compliance value did not account for increased upstream
GHG emissions. EPA requested comment on providing some type of
incentive, the appropriateness of both the zero grams/mile and vehicle
multiplier incentive mechanisms, and on any alternative approaches for
addressing advanced technology vehicle incentives. EPA received many
comments on these issues, which will be briefly summarized below.
Although some environmental organizations and State agencies
supported the principle of including some type of regulatory incentive
mechanism, almost all of their comments were opposed to the combination
of both the zero grams/mile compliance value and multipliers in the
higher end of the proposed range of 1.2 to 2.0. The California Air
Resources Board stated that the proposed credits ``are excessive'' and
the Union of Concerned Scientists stated that it ``strongly objects''
to the approach that lacks ``technical justification'' by not
``accounting for upstream emissions.'' The Natural Resources Defense
Council (NRDC) stated that the credits could ``undermine the emissions
benefits of the program and will have the unintended consequence of
slowing the development of conventional cleaner vehicle emission
reduction technologies into the fleet.'' NRDC, along with several other
commenters who made the same point, cited an example based on Nissan's
public statements that it plans on producing up to 150,000 Nissan Leaf
EVs in the near future at its plant in Smyrna, Tennessee.\231\ NRDC's
analysis showed that if EVs were to account for 10% of Nissan's car
fleet in 2016, the combination of the zero grams/mile and 2.0
multiplier would allow Nissan to make only relatively small
improvements to its gasoline car fleet and still be in compliance. NRDC
described a detailed methodology for calculating ``true full fuel cycle
emissions impacts'' for EVs. The Sierra Club suggested that the zero
grams/mile credit would ``taint'' EVs as the public comes to understand
that these vehicles are not zero-GHG vehicles, and that the zero grams/
mile incentive would allow higher gasoline vehicle GHG emissions.
---------------------------------------------------------------------------

\231\ ``Secretary Chu Announces Closing of $1.4 Billion Loan to
Nissan,'' Department of Energy, January 28, 2010, http://
www.energy.gov/news/8581.htm. EPA Docket EPA-HQ-OAR-2009-0472.
---------------------------------------------------------------------------

Most vehicle manufacturers were supportive of both the zero grams/
mile compliance value and a higher vehicle multiplier. The Alliance of
Automobile Manufacturers supported zero grams/mile ``since customers
need to receive a clear signal that they have made the right choice by
preferring an EV, PHEV, or EREV. * * * However, the Alliance recognizes
the need for a comprehensive approach with shared responsibility in
order to achieve an overall carbon reduction.'' Nissan claimed that
zero grams/mile is ``legally required,'' stating that EPA's 2-cycle
test procedures do not account for upstream GHG emissions, that
accounting for upstream emissions from electric vehicles but not from
other vehicles would be arbitrary, and that including upstream GHG
would ``disrupt the careful balancing embedded into the National
Program.'' Several other manufacturers, including Ford, Chrysler,
Toyota, and Mitsubishi, also supported the proposed zero grams/mile
compliance value. BMW suggested a compliance value approach similar to

[[Page 25436]]

that used for CAFE compliance (described below), which would yield a
very low, non-zero grams/mile compliance value. Honda opposed the zero
grams/mile incentive. Honda suggested that EPA should fully account for
upstream GHG and ``should separate incentives and credits from the
measurement of emissions.'' Automakers universally supported higher
multipliers, many higher than the maximum 2.0 level proposed by EPA.
Honda suggested a multiplier of 16.0 for FCVs. Mitsubishi supported the
concept of larger, temporary incentives until advanced technology
vehicle sales achieved a 10% market share. Finally, some commenters
suggested that other technologies should also receive incentives, such
as diesel vehicles, hydrogen-fueled internal combustion engines, and
natural gas vehicles.
Based on a careful consideration of these comments, EPA is
modifying its proposed advanced technology vehicle incentive program
for EVs, PHEVs, and FCVs produced in 2012-2016. EPA is not extending
the program to include additional technologies at this time. The final
incentive program, and our rationale for it, are described below.
One, the incentive program retains the zero grams/mile value for
EVs and FCVs, and for PHEVs when operated on grid electricity, subject
to vehicle production caps discussed below. EPA acknowledges that,
based on current electricity and hydrogen production processes, that
EVs, PHEVs, and FCVs yield higher upstream GHG emissions than
comparable gasoline vehicles. But EPA reiterates its support for
temporarily rewarding advanced emissions control technologies by
foregoing modest emissions reductions in the short term in order to lay
the foundation for the potential for much larger emission reductions in
the longer term.\232\ EPA notes that EVs, PHEVs, and FCVs are potential
GHG ``game changers'' if major cost and consumer barriers can be
overcome and if there is a nationwide transformation to low-GHG
electricity (or hydrogen, in the case of FCVs).
---------------------------------------------------------------------------

\232\ EPA has adopted this strategy in several of its most
recent and important mobile source rulemakings, such as its Tier 2
Light-Duty Vehicle, 2007 Heavy-Duty Highway, and Tier 4 Nonroad
Diesel rulemakings.
---------------------------------------------------------------------------

Although EVs and FCVs will have compliance values of zero grams/
mile, PHEV compliance values will be determined by combining zero
grams/mile for grid electricity operation with the GHG emissions from
the 2-cycle test results during operation on liquid fuel, and weighting
these values by the percentage of miles traveled that EPA believes will
be performed on grid electricity and on liquid fuel, which will vary
for different PHEVs. EPA is currently considering different approaches
for determining the weighting factor to be used in calculating PHEV GHG
emissions compliance values. EPA will consider the work of the Society
of Automotive Engineers Hybrid Technical Standards Committee, as well
as other relevant factors. EPA will issue a final rule on this
methodology by the fall of 2010, when EPA expects some PHEVs to
initially enter the market.
EPA agrees with the comments by the environmental organizations,
States, and Honda that the zero grams/mile compliance value will reduce
the overall GHG benefits of the program. However, EPA believes these
reductions in GHG benefits will be relatively small based on the
projected production of EVs, PHEVs, and FCVs during the 2012-2016
timeframe, along with the other changes that we are making in the
incentive program. EPA believes this modest potential for reduction in
near-term emissions control is more than offset by the potential for
very large future emissions reductions that commercialization of these
technologies could promote.
Two, the incentive program will not include any vehicle
multipliers, i.e., an EV's zero grams/mile compliance value will count
as one vehicle in a manufacturer's fleet average, not as more than one
vehicle as proposed. EPA has concluded that the combination of the zero
grams/mile and multiplier credits would be excessive. Compared to the
maximum multiplier of 2.0 that EPA had proposed, dropping this
multiplier reduces the aggregate impact of the overall credit program
by a factor of two (less so for lower multipliers, of course).
Three, EPA is placing a cumulative cap on the total production of
EVs, PHEVs, and FCVs for which an individual manufacturer can claim the
zero grams/mile compliance value during model years 2012-2016. The
cumulative production cap will be 200,000 vehicles, except those
manufacturers that sell at least 25,000 EVs, PHEVs, and FCVs in MY 2012
will have a cap of 300,000 vehicles for MY 2012-2016. This higher cap
option is an additional incentive for those manufacturers that take an
early leadership role in aggressively and successfully marketing
advanced technology vehicles. These caps are a second way to limit the
potential GHG benefit losses associated with the incentive program and
therefore are another response to the concerns that the proposed
incentives were excessive and could significantly undermine the
program's GHG benefits. If, for example, 500,000 EVs were produced in
2012-2016 that qualified for the zero grams/mile compliance value, the
loss in GHG benefits due to this program would be about 25 million
metric tons, or less than 3 percent of the total projected GHG benefits
of this program.\233\ The rationale for these caps is that the
incentive for EVs, PHEVs, and FCVs is most critical when individual
automakers are beginning to introduce advanced technologies in the
market, and less critical once individual automakers have successfully
achieved a reasonable market share and technology costs decline due to
higher production volumes and experience. EPA believes that cap levels
of 200,000-300,000 vehicles over a five model year period are
reasonable, as production greater than this would indicate that the
manufacturer has overcome at least some of the initial market barriers
to these advanced technologies. Further, EPA believes that it is
unlikely that many manufacturers will approach these cap levels in the
2012-2016 timeframe.\234\
---------------------------------------------------------------------------

\233\ See Regulatory Impact Analysis, Appendix 5.B. While it is,
of course, impossible to predict the number of EVs, PHEVs, and FCVs
that will be produced between 2012 and 2016 with absolute certainty,
EPA believes that 500,000 ``un-capped'' EVs is an optimistic
scenario. Fewer EVs, or a combination of 500,000 EVs and PHEVs,
would lessen the short-term reduction in GHG benefits. Production of
more than 500,000 ``un-capped'' EVs would increase the short-term
reduction in GHG benefits.
\234\ Fundamental power train changes in the automotive market
typically evolve slowly over time. For example, over ten years after
the U.S. introduction of the first conventional hybrid electric
vehicle, total hybrid sales are approximately 300,000 units per
year.
---------------------------------------------------------------------------

Production beyond the cumulative vehicle production cap for a given
manufacturer in MY 2012-2016 would have its compliance values
calculated according to a methodology that accounts in full for the net
increase in upstream GHG emissions. For an EV, for example, this would
involve: (1) Measuring the vehicle electricity consumption in watt-
hours/mile over the 2-cycle test (in the example introduced earlier, a
midsize EV might have a 2-cycle test electricity consumption of 230
watt-hours/mile), (2) adjusting this watt-hours/mile value upward to
account for electricity losses during transmission and vehicle charging
(dividing 230 watt-hours/mile by 0.93 to account for grid/transmission
losses and by 0.90 to reflect losses during vehicle charging yields a
value of 275 watt-hours/mile), (3) multiplying the adjusted watt-hours/
mile value by a

[[Page 25437]]

nationwide average electricity upstream GHG emissions rate of 0.642
grams/watt-hour at the powerplant \235\ (275 watt-hours/mile multiplied
by 0.642 grams GHG/watt-hour yields 177 grams/mile), and 4) subtracting
the upstream GHG emissions of a comparable midsize gasoline vehicle of
56 grams/mile to reflect a true net increase in upstream GHG emissions
(177 grams/mile for the EV minus 56 grams/mile for the gasoline vehicle
yields a net increase and EV compliance value of 121 grams/
mile).^{236 237} The full accounting methodology for the
portion of PHEV operation on grid electricity would use this same
approach.
---------------------------------------------------------------------------

\235\ The nationwide average electricity upstream GHG emissions
rate of 0.642 grams GHG/watt-hour was calculated from 2005
nationwide powerplant data for CO2, CH4, and
N2O emissions from eGRID2007 (http://www.epa.gov/
cleanenergy/energy-resources/egrid/index.html), converting to
CO2 -e using Global Warming Potentials of 25 for
CH4 and 298 for N2O, and multiplying by a
factor of 1.06 to account for GHG emissions associated with
feedstock extraction, transportation, and processing (based on
Argonne National Laboratory's The Greenhouse Gases, Regulated
Emissions, and Energy Use in Transportation (GREET) Model, Version
1.8c.0, available at http://www.transportation.anl.gov/modeling_
simulation/GREET/). EPA Docket EPA-HQ-OAR-2009-0472. EPA recognizes
that there are many issues involved with projecting the electricity
upstream GHG emissions associated with future EV and PHEV use
including, but not limited to, average vs marginal, daytime vs
nighttime vehicle charging, geographical differences, and changes in
future electricity feedstocks. EPA chose to use the 2005 national
average value because it is known and documentable. Values
appropriate for future vehicle use may be higher or lower than this
value. EPA will reevaluate this value in future rulemakings.
\236\ A midsize gasoline vehicle with a footprint of 45 square
feet would have a MY 2016 GHG target of about 225 grams/mile;
dividing 8887 grams CO2/gallon of gasoline by 225 grams/
mile yields an equivalent fuel economy level of 39.5 mpg; and
dividing 2208 grams upstream GHG/gallon of gasoline by 39.5 mpg
yields a midsize gasoline vehicle upstream GHG value of 56 grams/
mile. The 2208 grams upstream GHG/gallon of gasoline is calculated
from 19,200 grams upstream GHG/mmBtu (Renewable Fuel Standard
Program, Regulatory Impact Analysis, Section 2.5.8, February 2010)
and multiplying by 0.115 mmBtu/gallon of gasoline.
\237\ Manufacturers can utilize alternate calculation
methodologies if shown to yield equivalent or superior results and
if approved in advance by the Administrator.
---------------------------------------------------------------------------

EPA projects that the aggregate impact of the incentive program on
advanced technology vehicle GHG compliance values will be similar to
the way advanced technologies are treated under DOT's CAFE program. In
the CAFE program, the mpg value for an EV is determined using a
``petroleum equivalency factor'' that has a 1/0.15 factor built into it
similar to the flexible fuel vehicle credit.\238\ For example, under
current regulations, an EV with a 2-cycle electricity consumption of
230 watt-hours/mile would have a CAFE rating of about 360 miles per
gallon, which would be equivalent to a gasoline vehicle GHG emissions
value of 25 grams/mile, which is close to EPA's zero grams/mile for EV
production that is below an individual automaker's cumulative vehicle
production cap. The exception would be if a manufacturer exceeded its
cumulative vehicle production cap during MY 2012-2016. Then, the same
EV would have a GHG compliance value of about 120 grams/mile, which
would be significantly higher than the 25 gram/mile implied by the 360
mile/gallon CAFE value.
---------------------------------------------------------------------------

\238\ 65 FR 36987 (June 12, 2000).
---------------------------------------------------------------------------

EPA disagrees with Nissan that excluding upstream GHGs is legally
required under section 202(a)(1). In this rulemaking, EPA is adopting
standards under section 202(a)(1), which provides EPA with broad
discretion in setting emissions standards. This includes authority to
structure the emissions standards in a way that provides an incentive
to promote advances in emissions control technology. This discretion
includes the adjustments to compliance values adopted in the final
rule, the multipliers we proposed, and other kinds of incentives. EPA
recognizes that we have not previously made adjustments to a compliance
value to account for upstream emissions in a section 202(a) vehicle
emissions standard, but that does not mean we do not have authority to
do so in this case. In addition, EPA is not directly regulating
upstream GHG emissions from stationary sources, but instead is deciding
how much value to assign to a motor vehicle for purposes of compliance
calculations with the motor vehicle standard. While the logical place
to start is the emissions level measured under the test procedure,
section 202(a)(1) does not require that EPA limit itself to only that
level. For vehicles above the production volume cap described above,
EPA will adjust the measured value to a level that reflects the net
difference in upstream GHG emissions compared to a comparable
conventional vehicle. This will account for the actual GHG emissions
increase associated with the use of the EV. As shown above, upstream
GHG emissions attributable to increased electricity production to
operate EVs or PHEVs currently exceed the upstream GHG emissions
attributable to gasoline vehicles. There is a rational basis for EPA to
account for this net difference, as that best reflects the real world
effect on the air pollution problem we are addressing. For vehicles
above the cap, EPA is reasonably and fairly accounting for the
incremental increase in upstream GHG emissions from both the electric
vehicles and the conventional vehicles. EPA is not, as Nissan
suggested, arbitrarily counting upstream emissions for electric
vehicles but not for conventional fuel vehicles.
EPA recognizes that every motor vehicle fuel and fuel production
process has unique upstream GHG emissions impacts. EPA has discretion
in this rulemaking under section 202(a) on whether to account for
differences in net upstream GHG emissions relative to gasoline produced
from oil, and intends to only consider upstream GHG emissions for those
fuels that have significantly higher or lower GHG emissions impacts. At
this time, EPA is only making such a determination for electricity,
given that, as shown above in the example for a midsize car,
electricity upstream GHG emissions are about three times higher than
gasoline upstream GHG emissions. For example, the difference in
upstream GHG emissions for both diesel fuel from oil and CNG from
natural gas are relatively small compared to differences associated
with electricity. Nor is EPA arbitrarily ignoring upstream GHG
emissions of flexible fuel vehicles (FFVs) that can operate on E85.
Data show that, on average, FFVs operate on gasoline over 99 percent of
the time, and on E85 fuel less than 1 percent of the time.\239\ EPA's
recently promulgated Renewable Fuel Standard Program shows that, with
respect to aggregate lifecycle emissions including non-tailpipe GHG
emissions (such as feedstock growth, transportation, fuel production,
and land use), lifecycle emissions for ethanol from corn using advanced
production technologies are about 20 percent less GHG than gasoline
from oil.\240\ Given this difference, and that E85 is used in FFVs less
than 1 percent of the time, EPA has concluded that it is not necessary
to adopt a more complicated upstream accounting for FFVs. Accordingly,
EPA's incentive approach here is both reasonable and authorized under
section 202(a)(1).
---------------------------------------------------------------------------

\239\ Renewable Fuel Standard Program (RFS2), Regulatory Impact
Analysis, Section 1.7.4, February 2010.
\240\ 75 FR 14670 (March 26, 2010).
---------------------------------------------------------------------------

In summary, EPA believes that this program for MY 2012-2016 strikes
a reasoned balance by providing a temporary regulatory incentive to
help promote commercialization of advanced vehicle technologies which
are potential game-changers, but which also face major barriers, while
effectively minimizing potential GHG losses by dropping the proposed
multiplier and adding individual automaker

[[Page 25438]]

production volume caps. In the future, if there were a program to
control utility GHG emissions, then these advanced technology vehicles
have the potential to produce very large reductions in GHG emissions,
and to transform the transportation sector's contribution to nationwide
GHG emissions. EPA will reassess the issue of how to address EVs,
PHEVs, and FCVs in rulemakings for model years 2017 and beyond based on
the status of advanced vehicle technology commercialization, the status
of upstream GHG control programs, and other relevant factors.
Finally, the criteria and definitions for what vehicles qualify for
the advanced technology vehicle incentives are provided in Section
III.E. These definitions for EVs, PHEVs, and FCVs ensure that only
credible advanced technology vehicles are provided the incentives.
4. Off-Cycle Technology Credits
As proposed, EPA is adopting an optional credit opportunity
intended to apply to new and innovative technologies that reduce
vehicle CO2 emissions, but for which the CO2
reduction benefits are not significantly captured over the 2-cycle test
procedure used to determine compliance with the fleet average standards
(i.e., ``off-cycle'').\241\ Eligible innovative technologies are those
that are relatively newly introduced in one or more vehicle models, but
that are not yet implemented in widespread use in the light-duty fleet.
EPA will not approve credits for technologies that are not innovative
or do not provide novel approaches to reducing greenhouse gas
emissions. Manufacturers must obtain EPA approval for new and
innovative technologies at the time of vehicle certification in order
to earn credits for these technologies at the end of the model year.
This approval must include the testing methodology to be used for
quantifying credits. Further, any credits for these off-cycle
technologies must be based on real-world GHG reductions not
significantly captured on the current 2-cycle tests and verifiable test
methods, and represent average U.S. driving conditions.
---------------------------------------------------------------------------

\241\ See final regulations at 40 CFR 86.1866-12(d).
---------------------------------------------------------------------------

Similar to the technologies used to reduce A/C system indirect
CO2 emissions by increasing A/C efficiency, eligible
technologies would not be primarily active during the 2-cycle test and
therefore the associated improvements in CO2 emissions would
not be significantly captured. Because these technologies are not
nearly so well developed and understood, EPA is not prepared to
consider them in assessing the stringency of the CO2
standards. However, EPA is aware of some emerging and innovative
technologies and concepts in various stages of development with
CO2 reduction potential that might not be adequately
captured on the FTP or HFET. EPA believes that manufacturers should be
able to generate credit for the emission reductions these technologies
actually achieve, assuming these reductions can be adequately
demonstrated and verified. Examples include solar panels on hybrids or
electric vehicles, adaptive cruise control, and active aerodynamics.
EPA believes it would be appropriate to provide an incentive to
encourage the introduction of these types of technologies, that bona
fide reductions from these technologies should be considered in
determining a manufacturer's fleet average, and that a credit mechanism
is an effective way to do this. This optional credit opportunity would
be available through the 2016 model year.
EPA received comments from a few manufacturers that the ``new and
innovative'' criteria should be broadened. The commenters pointed out
that there are technologies already in the marketplace that would
provide emissions reductions off-cycle and that their use should be
incentivized. One manufacturer suggested that off-cycle credits should
be given for start-stop technologies. EPA does not agree that this
technology, which EPA's modeling projects will be widely used by
manufacturers in meeting the CO2 standards, should qualify
for off-cycle credits. Start-stop technology already achieves a
significant CO2 benefit on the current 2-cycle tests, which
is why many manufacturers have announced plans to adopt it across large
segments of the fleet. EPA recognizes there may be additional benefits
to start-stop technology beyond the 2-cycle tests (e.g., heavy idle
use), and that this is likely the case for other technologies that
manufacturers will rely on to meet the MY 2012-2016 standards. EPA
plans to continue to assess the off-cycle potential for these
technologies in the future. However, EPA does not believe that off-
cycle credits should be granted for technologies which we expect
manufacturers to rely on in widespread use throughout the fleet in
meeting the CO2 standards. Such credits could lead to double
counting, as there is already significant CO2 benefit over
the 2-cycle tests. EPA expects that most if not all technologies that
reduce CO2 emission on the 2-cycle test will also reduce
CO2 emissions during the wide variety of in-use operation
that is not directly captured in the 2-cycle test. This is no different
than what occurs from the control technology on vehicles for criteria
pollutants. We expect that the catalytic converter and other emission
control technology will operate to reduce emissions throughout in-use
driving, and not just when the vehicle is tested on the specified test
procedure. The aim for this off-cycle credit provisions is to provide
an incentive for technologies that normally would not be chosen as a
GHG control strategy, as their GHG benefits are not measured on the
specified 2-cycle test. It is not designed to provide credits for
technology that does provide significant GHG benefits on the 2-cycle
test and as expected will also typically provide GHG benefits in other
kinds of operation. Thus, EPA is finalizing the ``new and innovative''
criteria as proposed. That is, the potential to earn off-cycle credits
will be limited to those technologies that are new and innovative, are
introduced in only a limited number of vehicle models (i.e., not in
widespread use), and are not captured on the current 2-cycle tests.
This approach will encourage future innovation, which may lead to the
opportunity for future emissions reductions.
As proposed, manufacturers would quantify CO2 reductions
associated with the use of the innovative off-cycle technologies such
that the credits could be applied on a g/mile equivalent basis, as is
the case with A/C system improvements. Credits must be based on real
additional reductions of CO2 emissions and must be
quantifiable and verifiable with a repeatable methodology. As proposed,
the technologies upon which the credits are based would be subject to
full useful life compliance provisions, as with other emissions
controls. Unless the manufacturer can demonstrate that the technology
would not be subject to in-use deterioration over the useful life of
the vehicle, the manufacturer must account for deterioration in the
estimation of the credits in order to ensure that the credits are based
on real in-use emissions reductions over the life of the vehicle.
As discussed below, EPA is finalizing a two-tiered process for
demonstrating the CO2 reductions of an innovative and novel
technology with benefits not captured by the FTP and HFET test
procedures. First, a manufacturer must determine whether the benefit of
the technology could be captured using the 5-cycle methodology
currently used to determine fuel economy label values. EPA established
the 5-cycle test

[[Page 25439]]

methods to better represent real-world factors impacting fuel economy,
including higher speeds and more aggressive driving, colder temperature
operation, and the use of air conditioning. If this determination is
affirmative, the manufacturer must follow the procedures described
below (as codified in today's rules). If the manufacturer finds that
the technology is such that the benefit is not adequately captured
using the 5-cycle approach, then the manufacturer would have to develop
a robust methodology, subject to EPA approval, to demonstrate the
benefit and determine the appropriate CO2 gram per mile
credit. As discussed below, EPA is also providing opportunity for
public comment as part of the approval process for such non-5-cycle
credits.
a. Technology Demonstration Using EPA 5-Cycle Methodology
As noted above, the CO2 reduction benefit of some
innovative technologies could be demonstrated using the 5-cycle
approach currently used for EPA's fuel economy labeling program. The 5-
cycle methodology was finalized in EPA's 2006 fuel economy labeling
rule,\242\ which provides a more accurate fuel economy label estimate
to consumers starting with 2008 model year vehicles. In addition to the
FTP and HFET test procedures, the 5-cycle approach folds in the test
results from three additional test procedures to determine fuel
economy. The additional test cycles include cold temperature operation,
high temperature, high humidity and solar loading, and aggressive and
high-speed driving; thus these tests could be used to demonstrate the
benefit of a technology that reduces CO2 over these types of
driving and environmental conditions. Using the test results from these
additional test cycles collectively with the 2-cycle data provides a
more precise estimate of the average fuel economy and CO2
emissions of a vehicle for both the city and highway independently. A
significant benefit of using the 5-cycle methodology to measure and
quantify the CO2 reductions is that the test cycles are
properly weighted for the expected average U.S. operation, meaning that
the test results could be used without further adjustments.
---------------------------------------------------------------------------

\242\ Fuel Economy Labeling of Motor Vehicles: Revisions to
Improve Calculation of Fuel Economy Estimates; Final Rule (71 FR
77872, December 27, 2006).
---------------------------------------------------------------------------

EPA continues to believe that the use of these supplemental cycles
may provide a method by which technologies not demonstrated on the
baseline 2-cycles can be quantified and is finalizing this approach as
proposed. The cold temperature FTP can capture new technologies that
improve the CO2 performance of vehicles during colder
weather operation. These improvements may be related to warm-up of the
engine or other operation during the colder temperature. An example of
such a new, innovative technology is a waste heat capture device that
provides heat to the cabin interior, enabling additional engine-off
operation during colder weather not previously enabled due to heating
and defrosting requirements. The additional engine-off time would
result in additional CO2 reductions that otherwise would not
have been realized without the heat capture technology.
Although A/C credits for efficiency improvements will largely be
captured in the A/C credits provisions through the credit menu of known
efficiency improving components and controls, certain new technologies
may be able to use the high temperatures, humidity, and solar load of
the SC03 test cycle to accurately measure their impact. An example of a
new technology may be a refrigerant storage device that accumulates
pressurized refrigerant during driving operation or uses recovered
vehicle kinetic energy during deceleration to pressurize the
refrigerant. Much like the waste heat capture device used in cold
weather, this device would also allow additional engine-off operation
while maintaining appropriate vehicle interior occupant comfort levels.
SC03 test data measuring the relative impact of innovative A/C-related
technologies could be applied to the 5-cycle equation to quantify the
CO2 reductions of the technology.
The US06 cycle may be used to capture innovative technologies
designed to reduce CO2 emissions during higher speed and
more aggressive acceleration conditions, but not reflected on the 2-
cycle tests. An example of this is an active aerodynamic technology.
This technology recognizes the benefits of reduced aerodynamic drag at
higher speeds and makes changes to the vehicle at those speeds. The
changes may include active front or grill air deflection devices
designed to redirect frontal airflow. Certain active suspension devices
designed primarily to reduce aerodynamic drag by lowering the vehicle
at higher speeds may also be measured on the US06 cycle. To properly
measure these technologies on the US06, the vehicle would require
unique load coefficients with and without the technologies. The
different load coefficient (properly weighted for the US06 cycle) could
effectively result in reduced vehicle loads at the higher speeds when
the technologies are active. Similar to the previously discussed
cycles, the results from the US06 test with and without the technology
could then use the 5-cycle methodology to quantify CO2
reductions.
If the 5-cycle procedures can be used to demonstrate the innovative
technology, then the regulatory evaluation/approval process will be
relatively simple. The manufacturer will simply test vehicles with and
without the technology installed or operating and compare results. All
5-cycles must be tested with the technology enabled and disabled, and
the test results will be used to calculate a combined city/highway
CO2 value with the technology and without the technology.
These values will then be compared to determine the amount of the
credit; the combined city/highway CO2 value with the
technology operating will be subtracted from the combined city/highway
CO2 value without the technology operating to determine the
gram per mile CO2 credit. It is likely that multiple tests
of each of the five test procedures will need to be performed in order
to achieve the necessary strong degree of statistical significance of
the credit determination results. This will have to be done for each
model type for which a credit is sought, unless the manufacturer could
demonstrate that the impact of the technology was independent of the
vehicle configuration on which it was installed. In this case, EPA may
consider allowing the test to be performed on an engine family basis or
other grouping. At the end of the model year, the manufacturer will
determine the number of vehicles produced subject to each credit amount
and report that to EPA in the final model year report. The gram per
mile credit value determined with the 5-cycle comparison testing will
be multiplied by the total production of vehicles subject to that value
to determine the total number of credits.
EPA received a few comments regarding the 5-cycle approach. While
not commenting directly on the 5-cycle testing methodology, the
Alliance raised general concerns that the proposed approach did not
offer manufacturers enough certainty with regard to credit applications
and testing in order to take advantage of the credits. The Alliance
further commented that the proposal did not provide a level playing
field to all manufacturers in terms of possible credit availability.
The Alliance recommended that rather than attempting to quantify
CO2 reductions with a prescribed test procedure on unknown
technologies, EPA should

[[Page 25440]]

handle credit applications and testing guidelines via future guidance
letters, as technologies emerge and are developed.
EPA believes that 5-cycle testing methodology is one clear and
objective way to demonstrate certain off-cycle emissions control
technologies, as discussed above. It provides certainty with regard to
testing, and is available for all manufacturers. As discussed below,
there are also other options for manufactures where the 5-cycle test is
not appropriate. EPA is retaining this as a primary methodology for
determining off-cycle credits. For technologies not able to be
demonstrated on the 5-cycle test, EPA is finalizing an approach that
will include a public comment opportunity, as discussed below, which we
believe addresses commenter concerns regarding maintaining a level
playing field.
b. Alternative Off-Cycle Credit Methodologies
As proposed, in cases where the benefit of a technological approach
to reducing CO2 emissions can not be adequately represented
using existing test cycles, manufacturers will need to develop test
procedures and analytical approaches to estimate the effectiveness of
the technology for the purpose of generating credits. As discussed
above, the first step must be a thorough assessment of whether the 5-
cycle approach can be used to demonstrate a reduction in emissions. If
EPA determines that the 5-cycle process is inadequate for the specific
technology being considered by the manufacturer (i.e., the 5-cycle test
does not demonstrate any emissions reductions), then an alternative
approach may be developed and submitted to EPA for approval. The
demonstration program must be robust, verifiable, and capable of
demonstrating the real-world emissions benefit of the technology with
strong statistical significance.
The CO2 benefit of some technologies may be able to be
demonstrated with a modeling approach, using engineering principles. An
example would be where a roof solar panel is used to charge the on-
board vehicle battery. The amount of potential electrical power that
the panel could supply could be modeled for average U.S. conditions and
the units of electrical power could be translated to equivalent fuel
energy or annualized CO2 emission rate reduction from the
captured solar energy. The CO2 reductions from other
technologies may be more challenging to quantify, especially if they
are interactive with the driver, geographic location, environmental
condition, or other aspect related to operation on actual roads. In
these cases, manufacturers might have to design extensive on-road test
programs. Any such on-road testing programs would need to be
statistically robust and based on average U.S. driving conditions,
factoring in differences in geography, climate, and driving behavior
across the U.S.
Whether the approach involves on-road testing, modeling, or some
other analytical approach, the manufacturer will be required to present
a proposed methodology to EPA. EPA will approve the methodology and
credits only if certain criteria are met. Baseline emissions and
control emissions must be clearly demonstrated over a wide range of
real world driving conditions and over a sufficient number of vehicles
to address issues of uncertainty with the data. The analytical approach
must be robust, verifiable, and capable of demonstrating the real-world
emissions benefit with strong statistical significance. Data must be on
a vehicle model-specific basis unless a manufacturer demonstrated model
specific data was not necessary. Approval of the approach to
determining a CO2 benefit will not imply approval of the
results of the program or methodology; when the testing, modeling, or
analyses are complete the results will likewise be subject to EPA
review and approval. EPA believes that manufacturers could work
together to develop testing, modeling, or analytical methods for
certain technologies, similar to the SAE approach used for A/C
refrigerant leakage credits.
In addition, EPA received several comments recommending that the
approval process include an opportunity for public comment. As noted
above, some manufacturers are concerned that there be a level playing
field in terms of all manufacturers having a reasonable opportunity to
earn credits under an approved approach. Commenters also want an
opportunity for input in the methodology to ensure the accuracy of
credit determinations for these technologies. Commenters point out that
there are a broad number of stakeholders with experience in the issues
pertaining to the technologies that could add value in determining the
most appropriate method to assess these technologies' performance. EPA
agrees with these comments and is including an opportunity for public
comment as part of the approval process. If and when EPA receives an
application for off-cycle credits using an alternative non 5-cycle
methodology, EPA will publish a notice of availability in the Federal
Register with instructions on how to comment on draft off-cycle credit
methodology. The public information available for review will focus on
the methodology for determining credits but the public review obviously
is limited to non-confidential business information. The timing for
final approval will depend on the comments received. EPA also believes
that a public review will encourage manufacturers to be thorough in
their preparation prior to submitting their application for credits to
EPA for approval. EPA will take comments into consideration, and where
appropriate, work with the manufacturer to modify their approach prior
to approving any off-cycle credits methodology. EPA will give final
notice of its determination to the general public as well as the
applicant. Off-cycle credits would be available in the model year
following the final approval. Thus, it will be imperative for a
manufacturer pursuing this option to begin the process as early as
possible.
EPA also received comments that the off-cycle credits highlights
the inadequacy of current test procedures, and that there is a clear
need for updated certification test procedures. As discussed in Section
III. B., EPA believes the current test procedures are adequate for
implementing the standards finalized today. However, EPA is interested
in improving test procedures in the future and believes that the off-
cycle credits program has the potential to provide useful data and
insights both for the 5-cycle test procedures and also other test
procedures that capture off-cycle emissions.
5. Early Credit Options
EPA is finalizing a program to allow manufacturers to generate
early credits in model years 2009-2011.\243\ As described below,
credits may be generated through early additional fleet average
CO2 reductions, early A/C system improvements, early
advanced technology vehicle credits, and early off-cycle credits. As
with other credits, early credits are subject to a five year carry-
forward limit based on the model year in which they are generated.
Manufacturers may transfer early credits between vehicle categories
(e.g., between the car and truck fleet). With the exception of MY 2009
early program credits, as discussed below, a manufacturer may trade
other early credits to other manufacturers without limits. The agencies
note that CAFE credits earned in MYs prior to MY 2011 will still be
available to manufacturers

[[Page 25441]]

for use in the CAFE program in accordance with applicable regulations.
---------------------------------------------------------------------------

\243\ See final regulations at 40 CFR 86.1867-12.
---------------------------------------------------------------------------

EPA is not adopting certification, compliance, or in-use
requirements for vehicles generating early credits. Since manufacturers
are already certifying MY 2010 and in some cases even MY 2011 vehicles,
doing so would make certification, compliance, and in-use requirements
unworkable. As discussed below, manufacturers are required to submit an
early credits report to EPA for approval no later than 90 days after
the end of MY 2011. This report must include details on all early
credits the manufacturer generates, why the credits are bona fide, how
they are quantified, and how they can be verified.
a. Credits Based on Early Fleet Average CO2 Reductions
As proposed, EPA is finalizing opportunities for early credit
generation in MYs 2009-2011 through over-compliance with a fleet
average CO2 baseline established by EPA. EPA is finalizing
four pathways for doing so. In order to generate early CO2
credits, manufacturers must select one of the four paths for credit
generation for the entire three year period and may not switch between
pathways for different model years. For two pathways, EPA is
establishing the baseline equivalent to the California standards for
the relevant model year. Generally, manufacturers that over-comply with
those CARB standards would earn credits. Two additional pathways,
described below, include credits based on over-compliance with CAFE
standards in states that have not adopted the California standards.
EPA received comments from manufacturers in support of the early
credits program as a necessary compliance flexibility. The Alliance
commented that the early credits reward manufacturers for providing
fleet performance that exceeds California and Federal standards and do
not result in a windfall. AIAM commented that early credits are
essential to assure the feasibility of the proposed standards and the
need for such credits must be evaluated in the context of the dramatic
changes the standards will necessitate in vehicle design and the
current economic environment in which manufacturers are called upon to
make the changes. Manufacturers also supported retaining all four
pathways, commenting that eliminating pathways would diminish the
flexibility of the program. EPA also received comments from many
environmental organizations and states that the program would provide
manufacturers with windfall credits because manufacturers will not have
to take any steps to earn credits beyond those that are already planned
and in some cases implemented. These commenters were particularly
concerned that the California truck standards in MY 2009 are not as
stringent as CAFE, so overcompliance with the California standards
could be a windfall in MY 2009, and possibly even MY 2010. These
commenters supported an early credits program based on overcompliance
with the more stringent of either the CAFE or California standards in
any given year. EPA is retaining the early credits program because EPA
judges that they are not windfall credits, and manufacturers in some
cases have reasonably relied on the availability of these credits, and
have based early model year compliance strategies on their availability
so that the credits are needed to provide adequate lead for the initial
years of the program. However, as discussed below, EPA is restricting
credit trading for MY 2009 credits earned under the California-based
pathways.
Manufacturers selecting Pathway 1 will generate credits by over-
complying with the California equivalent baseline established by EPA
over the manufacturer's fleet of vehicles sold nationwide.
Manufacturers selecting Pathway 2 will generate credits against the
California equivalent baseline only for the fleet of vehicles sold in
California and the CAA section 177 states.\244\ This approach includes
all CAA 177 states as of the date of promulgation of the Final Rule in
this proceeding. Manufacturers are required to include both cars and
trucks in the program. Under Pathways 1 and 2, EPA is requiring
manufacturers to cover any deficits incurred against the baseline
levels established by EPA during the three year period 2009-2011 before
credits can be carried forward into the 2012 model year. For example, a
deficit in 2011 would have to be subtracted from the sum of credits
earned in 2009 and 2010 before any credits could be applied to 2012 (or
later) model year fleets. EPA is including this provision to help
ensure the early credits generated under this program are consistent
with the credits available under the California program during these
model years. In its comments, California supported such an approach.
---------------------------------------------------------------------------

\244\ CAA 177 states refers to states that have adopted the
California GHG standards. At present, there are thirteen CAA 177
states: New York, Massachusetts, Maryland, Vermont, Maine,
Connecticut, Arizona, New Jersey, New Mexico, Oregon, Pennsylvania,
Rhode Island, Washington, as well as Washington, DC.
---------------------------------------------------------------------------

Table III.C.5-1 provides the California equivalent baselines EPA is
finalizing to be used as the basis for CO2 credit generation
under the California-based pathways. These are the California GHG
standards for the model years shown. EPA proposed to adjust the
California standards by 2.0 g/mile to account for the exclusion of
N2O and CH4, which are included in the California
GHG standards, but not included in the credits program. EPA received
comments from one manufacturer that this adjustment is in error and
should not be made. The commenter noted that EPA already includes total
hydrocarbons in the carbon balance determination of carbon related
exhaust emissions and therefore already accounts for CH4.
EPA also includes CO in the carbon related exhaust emissions
determination which acts to offset the need for an N20
adjustment. The commenter noted that THC and CO add about 0.8 to 3.0 g/
mile to the determination of carbon related emissions and therefore EPA
should not make the 2.0g/mile adjustment. The commenter is correct, and
therefore the final levels shown in the table below are 2.0 g/mile
higher than proposed. These comments are further discussed in the
Response to Comments document. Manufacturers will generate
CO2 credits by achieving fleet average CO2 levels
below these baselines. As shown in the table, the California-based
early credit pathways are based on the California vehicle categories.
Also, the California-based baseline levels are not footprint-based, but
universal levels that all manufacturers would use. Manufacturers will
need to achieve fleet levels below those shown in the table in order to
earn credits, using the California vehicle category definitions.

[[Page 25442]]

Table III.C.5-1--California Equivalent Baselines CO2 Emissions Levels for Early Credit Generation
----------------------------------------------------------------------------------------------------------------
Light trucks with a LVW of
Passenger cars and light 3,751 or more and a GVWR
Model year trucks with an LVW of 0- of up to 8,500 lbs plus
3,750 lbs medium-duty passenger
vehicles
----------------------------------------------------------------------------------------------------------------
2009.................................................. 323 439
2010.................................................. 301 420
2011.................................................. 267 390
----------------------------------------------------------------------------------------------------------------

Manufacturers using Pathways 1 or 2 above will use year-end car and
truck sales in each category. Although production data is used for the
program starting in 2012, EPA is using sales data for the early credits
program in order to apportion vehicles by State. This is described
further below. Manufacturers must calculate actual fleet average
emissions over the appropriate vehicle fleet, either for vehicles sold
nationwide for Pathway 1, or California plus 177 states sales for
Pathway 2. Early CO2 credits are based on the difference
between the baseline shown in the table above and the actual fleet
average emissions level achieved. Any early A/C credits generated by
the manufacturer, described below in Section III.C.5.b, will be
included in the fleet average level determination. In model year 2009,
the California CO2 standard for cars (323 g/mi
CO2) is equivalent to 323 g/mi CO2, and the
California light-truck standard (437 g/mi CO2) is less
stringent than the equivalent CAFE standard, recognizing that there are
some differences between the way the California program and the CAFE
program categorize vehicles. Manufacturers are required to show that
they over comply over the entire three model year time period, not just
the 2009 model year, to generate early credits under either Pathways 1,
2 or 3. A manufacturer cannot use credits generated in model year 2009
unless they offset any debits from model years 2010 and 2011.
EPA received comments that this approach will provide windfall
credits to manufacturers because the MY 2009 California light truck
standards are less stringent than the corresponding CAFE standards.
While this could be accurate if credits were based on performance in
just MY 2009, that is not how credits are determined. Credits are based
on the performance over a three model year period, MY 2009-2011. As
noted in the proposal, EPA expects that the requirement to over comply
over the entire time period covering these three model years should
mean that the credits that are generated are real and are in excess of
what would have otherwise occurred. However, because of the
circumstances involving the 2009 model year, in particular for
companies with significant truck sales, there is some concern that
under Pathways 1, 2, and 3, there is a potential for a large number of
credits generated in 2009 against the California standard, in
particular for a number of companies who have significantly over-
achieved on CAFE in recent model years. Some commenters were very
concerned about this issue and commented in support of restricting
credit trading between firms of MY 2009 credits based on the California
program. EPA requested comments on this approach and is finalizing this
credit trading restriction based on continued concerns regarding the
issue of windfall credits. EPA wants to avoid a situation where,
contrary to expectation, some part of the early credits generated by a
manufacturer are in fact not excess, where companies could trade such
credits to other manufacturers, risking a delay in the addition of new
technology across the industry from the 2012 and later EPA
CO2 standards. Therefore, manufacturers selecting Pathways
1, 2, or 3 will not be allowed to trade any MY 2009 credits that they
may generate.
Commenters also recommended basing credits on the more stringent of
the standards between CAFE and CARB, which for MY 2009, would be the
CAFE standards. However, EPA believes that this would not be necessary
in light of the credit provisions requiring manufacturers choosing the
California based pathways to use the California pathway for all three
MYs 2009-2011, and the credit trading restrictions for MY 2009
discussed above.
In addition, for Pathways 1 and 2, EPA is allowing manufacturers to
include alternative compliance credits earned per the California
alternative compliance program.\245\ These alternative compliance
credits are based on the demonstrated use of alternative fuels in flex
fuel vehicles. As with the California program, the credits are
available beginning in MY 2010. Therefore, these early alternative
compliance credits are available under EPA's program for the 2010 and
2011 model years. FFVs are otherwise included in the early credit fleet
average based on their emissions on the conventional fuel. This does
not apply to EVs and PHEVs. The emissions of EVs and PHEVs are to be
determined as described in Section III.C.3. Manufacturers may choose to
either include their EVs and PHEVs in one of the four pathways
described in this section or under the early advanced technology
emissions credits described below, but not both due to issues of credit
double counting.
---------------------------------------------------------------------------

\245\ See Section 6.6.E, California Environmental Protection
Agency Air Resources Board, Staff Report: Initial Statement of
Reasons For Proposed Rulemaking, Public Hearing to Consider Adoption
of Regulations to Control Greenhouse Gas Emissions From Motor
Vehicles, August 6, 2004.
---------------------------------------------------------------------------

EPA is also finalizing two additional early credit pathways
manufacturers could select. Pathways 3 and 4 incorporate credits based
on over-compliance with CAFE standards for vehicles sold outside of
California and CAA 177 states in MY 2009-2011. Pathway 3 allows
manufacturers to earn credits as under Pathway 2, plus earn CAFE-based
credits in other states. Credits may not be generated for cars sold in
California and CAA 177 states unless vehicle fleets in those states are
performing better than the standards which otherwise would apply in
those states, i.e., the baselines shown in Table III.C.5-1 above.
Pathway 4 is for manufacturers choosing to forego California-based
early credits entirely and earn only CAFE-based credits outside of
California and CAA 177 states. Manufacturers may not include FFV
credits under the CAFE-based early credit pathways since those credits
do not automatically reflect actual reductions in CO2
emissions.
The baselines for CAFE-based early pathways are provided in Table
III.C.5-2 below. They are based on the CAFE standards for the 2009-2011
model years. For CAFE standards in 2009-2011 model years that are
footprint-based, the baseline would vary by manufacturer. Footprint-
based standards are in effect for the 2011 model year CAFE

[[Page 25443]]

standards.\246\ Additionally, for Reform CAFE truck standards,
footprint standards are optional for the 2009-2010 model years. Where
CAFE footprint-based standards are in effect, manufacturers will
calculate a baseline using the footprints and sales of vehicles outside
of California and CAA 177 states. The actual fleet CO2
performance calculation will also only include the vehicles sold
outside of California and CAA 177 states, and as mentioned above, may
not include FFV credits.
---------------------------------------------------------------------------

\246\ 74 FR 14196, March 30, 2009.

Table III.C.5-2--CAFE Equivalent Baselines CO2 Emissions Levels for
Early Credit Generation
------------------------------------------------------------------------
Model year Cars Trucks
------------------------------------------------------------------------
2009........................ 323................. 381 *
2010........................ 323................. 376 *
2011........................ Footprint-based Footprint-based
standard. standard.
------------------------------------------------------------------------
* Must be footprint-based standard for manufacturers selecting footprint
option under CAFE.

For the CAFE-based pathways, EPA is using the NHTSA car and truck
definitions that are in place for the model year in which credits are
being generated. EPA understands that the NHTSA definitions change
starting in the 2011 model year, and therefore changes part way through
the early credits program. EPA further recognizes that medium-duty
passenger vehicles (MDPVs) are not part of the CAFE program until the
2011 model year, and therefore are not part of the early credits
calculations for 2009-2010 under the CAFE-based pathways.
Pathways 2 through 4 involve splitting the vehicle fleet into two
groups, vehicles sold in California and CAA 177 states and vehicles
sold outside of these states. This approach requires a clear accounting
of location of vehicle sales by the manufacturer. EPA believes it will
be reasonable for manufacturers to accurately track sales by State,
based on its experience with the National Low Emissions Vehicle (NLEV)
Program. NLEV required manufacturers to meet separate fleet average
standards for vehicles sold in two different regions of the
country.\247\ As with NLEV, the determination is to be based on where
the completed vehicles are delivered as a point of first sale, which in
most cases would be the dealer.\248\
---------------------------------------------------------------------------

\247\ 62 FR 31211, June 6, 1997.
\248\ 62 FR 31212, June 6, 1997.
---------------------------------------------------------------------------

As noted above, manufacturers choosing to generate early
CO2 credits must select one of the four pathways for the
entire early credits program and would not be able to switch among
them. Manufacturers must submit their early credits report to EPA when
they submit their final CAFE report for MY 2011 (which is required to
be submitted no later than 90 days after the end of the model year).
Manufacturers will have until then to decide which pathway to select.
This gives manufacturers enough time to determine which pathway works
best for them. This timing may be necessary in cases where
manufacturers earn credits in MY 2011 and need time to assess data and
prepare an early credits submittal for final EPA approval.
The table below provides a summary of the four fleet average-based
CO2 early credit pathways EPA is finalizing:

Table III.C.5-3--Summary of Early Fleet Average CO2 Credit Pathways
------------------------------------------------------------------------

------------------------------------------------------------------------
Common Elements................... --Manufacturers select a pathway.
Once selected, may not switch among
pathways.
--All credits subject to 5 year
carry-forward restrictions.
--For Pathways 2-4, vehicles
apportioned by State based on point
of first sale.
Pathway 1: California-based --Manufacturers earn credits based
Credits for National Fleet. on fleet average emissions compared
with California equivalent baseline
set by EPA.
--Based on nationwide CO2 sales-
weighted fleet average.
--Based on use of California vehicle
categories.
--FFV alternative compliance credits
per California program may be
included.
--Once in the program, manufacturers
must make up any deficits that are
incurred prior to 2012 in order to
carry credits forward to 2012 and
later.
Pathway 2: California-based --Same as Pathway 1, but
Credits for vehicles sold in manufacturers only includes
California plus CAA 177 States. vehicles sold in California and CAA
177 states in the fleet average
calculation.
Pathway 3: Pathway 2 plus CAFE- --Manufacturer earns credits as
based Credits outside of provided by Pathway 2: California-
California plus CAA 177 States. based credits for vehicles sold in
California plus CAA 177 States,
plus:
--CAFE-based credits allowed for
vehicles sold outside of California
and CAA 177 states.
--For CAFE-based credits,
manufacturers earn credits based on
fleet average emissions compared
with baseline set by EPA.
--CAFE-based credits based on NHTSA
car and truck definitions.
--FFV credits not allowed to be
included for CAFE-based credits.
Pathway 4: Only CAFE-based Credits --Manufacturer elects to only earn
outside of California plus CAA CAFE-based credits for vehicles
177 States. sold outside of California and CAA
177 states. Earns no California and
177 State credits.
--For CAFE-based credits,
manufacturers earn credits based on
fleet average emissions compared
with baseline set by EPA.
--CAFE-based credits based on NHTSA
car and truck definitions.
--FFV credits not allowed to be
included for CAFE-based credits.
------------------------------------------------------------------------

[[Page 25444]]

b. Early A/C Credits
As proposed, EPA is finalizing provisions allowing manufacturers to
earn early A/C credits in MYs 2009-2011 using the same A/C system
design-based EPA provisions being finalized for MYs commencing in 2012,
as described in Section III.C.1, above. Manufacturers will be able to
earn early A/C CO2-equivalent credits by demonstrating
improved A/C system performance, for both direct and indirect
emissions. To earn credits for vehicles sold in California and CAA 177
states, the vehicles must be included in one of the California-based
early credit pathways described above in III.C.5.a. EPA is finalizing
this constraint in order to avoid credit double counting with the
California program in place in those states, which also allows A/C
system credits in this time frame. Manufacturers must fold the A/C
credits into the fleet average CO2 calculations under the
California-based pathway. For example, the MY 2009 California-based
program car baseline would be 323 g/mile (see Table III.C.5-1). If a
manufacturer under Pathway 1 had a MY 2009 car fleet average
CO2 level of 320 g/mile and then earned an additional 12 g/
mile CO2-equivalent A/C credit, the manufacturers would earn
a total of 10 g/mile of credit. Vehicles sold outside of California and
177 states would be eligible for the early A/C credits whether or not
the manufacturers participate in other aspects of the early credits
program. The early A/C credits for vehicles sold outside of California
and 177 states are based on the NHTSA vehicle categories established
for the model year in which early A/C credits are being earned.
c. Early Advanced Technology Vehicle Incentive
As discussed in Section III.C.3, above, EPA is finalizing an
incentive for sales of advanced technology vehicles including EVs,
PHEVs, and fuel cell vehicles. EPA is not including a multiplier for
these vehicles. However, EPA is allowing the use of the 0 g/mile value
for electricity operation for up to 200,000 vehicles per manufacturer
(or 300,000 vehicles for any manufacturer that sells 25,000 or more
advanced technology vehicles in MY 2012). EPA believes that providing
an incentive for the sales of such vehicles prior to MY 2012 is
consistent with the goal encouraging the introduction of such vehicles
as early as possible. Therefore, manufacturers may use the 0 g/mile
value for vehicles sold in MY 2009-2011 consistent with the approach
being finalized for MY 2012-2016. Any vehicles sold prior to MY 2012
under these provisions must be counted against the cumulative sales cap
of 200,000 (or 300,000, if applicable) vehicles. Manufacturers selling
such vehicles in MY 2009-2011 have the option of either folding them
into the early credits calculation under Pathways 1 through 4 described
in III.C.5.a above, or tracking the sales of these vehicles separately
for use in their fleetwide average compliance calculation in MY 2012 or
later years, but may not do both as this would lead to double counting.
Manufacturers tracking the sales of vehicles not folded into Pathways
1-4, may choose to use the vehicle counts along with the 0 g/mi
emissions value (up to the applicable vehicle sales cap) to comply with
2012 or later standards. For example, if a manufacturer sells 1,000 EVs
in MY 2011, the manufacturer would then be able to include 1,000
vehicles at 0 g/mile in their MY 2012 fleet to decrease the fleet
average for that model year. Again, these 1,000 vehicles would be
counted against the cumulative cap of 200,000 or 300,000, as
applicable, vehicles. Also, these 1,000 EVs would not be included in
the early credit pathways discussed above in Section III.C.5.a,
otherwise the vehicles would be double counted. As with early credits,
these early advanced technology vehicles will be tracked by model year
(2009, 2010, or 2011) and subject to the 5-year carry-forward
restrictions.
d. Early Off-Cycle Credits
EPA's is finalizing off-cycle innovative technology credit
provisions, as described in Section III.C.4. EPA requested comment on
beginning these credits in the 2009-2011 time frame, provided
manufacturers are able to make the necessary demonstrations outlined in
Section III.C.4, above. EPA is finalizing this approach for early off-
cycle credits as a way to encourage innovation to lower emissions as
early as possible, including the requirements for public review
described in Section III.C.4. Upon EPA approval of a manufacturer's
application for credits, the credits may be earned retroactively. EPA
did not receive comments specifically on early off-cycle credits.

D. Feasibility of the Final CO2 Standards

This final rule is based on the need to obtain significant GHG
emissions reductions from the transportation sector, and the
recognition that there are cost-effective technologies to achieve such
reductions for MY 2012-2016 vehicles. As in many prior mobile source
rulemakings, the decision on what standard to set is largely based on
the effectiveness of the emissions control technology, the cost and
other impacts of implementing the technology, and the lead time needed
for manufacturers to employ the control technology. The standards
derived from assessing these factors are also evaluated in terms of the
need for reductions of greenhouse gases, the degree of reductions
achieved by the standards, and the impacts of the standards in terms of
costs, quantified benefits, and other impacts of the standards. The
availability of technology to achieve reductions and the cost and other
aspects of this technology are therefore a central focus of this
rulemaking.
EPA is taking the same basic approach in this rulemaking, although
the technological problems and solutions involved in this rulemaking
differ in some ways from prior mobile source rulemakings. Here, the
focus of the emissions control technology is on reducing CO2
and other greenhouse gases. Vehicles combust fuel to perform two basic
functions: (1) To transport the vehicle, its passengers and its
contents (and any towed loads), and (2) to operate various accessories
during the operation of the vehicle such as the air conditioner.
Technology can reduce CO2 emissions by either making more
efficient use of the energy that is produced through combustion of the
fuel or reducing the energy needed to perform either of these
functions.
This focus on efficiency calls for looking at the vehicle as an
entire system, and the proposed and now final standards reflect this
basic paradigm. In addition to fuel delivery, combustion, and
aftertreatment technology, any aspect of the vehicle that affects the
need to produce energy must also be considered. For example, the
efficiency of the transmission system, which takes the energy produced
by the engine and transmits it to the wheels, and the resistance of the
tires to rolling both have major impacts on the amount of fuel that is
combusted while operating the vehicle. The braking system, the
aerodynamics of the vehicle, and the efficiency of accessories, such as
the air conditioner, all affect how much fuel is combusted as well.
In evaluating vehicle efficiency, we have excluded fundamental
changes in vehicles' size and utility. For example, we did not evaluate
converting minivans and SUVs to station wagons, converting vehicles
with four wheel drive to two wheel drive, or reducing headroom in order
to lower the roofline and reduce aerodynamic drag. We have

[[Page 25445]]

limited our assessment of technical feasibility and resultant vehicle
cost to technologies which maintain vehicle utility as much as
possible. Manufacturers may decide to alter the utility of the vehicles
which they sell in response to this rule, but this is not a necessary
consequence of the rule but rather a matter of automaker choice.
This need to focus on the efficient use of energy by the vehicle as
a system leads to a broad focus on a wide variety of technologies that
affect almost all the systems in the design of a vehicle. As discussed
below, there are many technologies that are currently available which
can reduce vehicle energy consumption. These technologies are already
being commercially utilized to a limited degree in the current light-
duty fleet. These technologies include hybrid technologies that use
higher efficiency electric motors as the power source in combination
with or instead of internal combustion engines. While already
commercialized, hybrid technology continues to be developed and offers
the potential for even greater efficiency improvements. Finally, there
are other advanced technologies under development, such as lean burn
gasoline engines, which offer the potential of improved energy
generation through improvements in the basic combustion process. In
addition, the available technologies are not limited to powertrain
improvements but also include mass reduction, electrical system
efficiencies, and aerodynamic improvements.
The large number of possible technologies to consider and the
breadth of vehicle systems that are affected mean that consideration of
the manufacturer's design and production process plays a major role in
developing the final standards. Vehicle manufacturers typically develop
many different models by basing them on a limited number of vehicle
platforms. The platform typically consists of a common set of vehicle
architecture and structural components. This allows for efficient use
of design and manufacturing resources. Given the very large investment
put into designing and producing each vehicle model, manufacturers
typically plan on a major redesign for the models approximately every 5
years. At the redesign stage, the manufacturer will upgrade or add all
of the technology and make most other changes supporting the
manufacturer's plans for the next several years, including plans
related to emissions, fuel economy, and safety regulations.
This redesign often involves a package of changes designed to work
together to meet the various requirements and plans for the model for
several model years after the redesign. This often involves significant
engineering, development, manufacturing, and marketing resources to
create a new product with multiple new features. In order to leverage
this significant upfront investment, manufacturers plan vehicle
redesigns with several model years' of production in mind. Vehicle
models are not completely static between redesigns as limited changes
are often incorporated for each model year. This interim process is
called a refresh of the vehicle and generally does not allow for major
technology changes although more minor ones can be done (e.g., small
aerodynamic improvements, valve timing improvements, etc.). More major
technology upgrades that affect multiple systems of the vehicle thus
occur at the vehicle redesign stage and not in the time period between
redesigns. The Center for Biological Diversity commented on EPA's
assumptions on redesign cycles, and these comments are addressed in
Section III.D.7 below.
As discussed below, there are a wide variety of CO2
reducing technologies involving several different systems in the
vehicle that are available for consideration. Many can involve major
changes to the vehicle, such as changes to the engine block and
cylinder heads, redesign of the transmission and its packaging in the
vehicle, changes in vehicle shape to improve aerodynamic efficiency and
the application of aluminum (and other lightweight materials) in body
panels to reduce mass. Logically, the incorporation of emissions
control technologies would be during the periodic redesign process.
This approach would allow manufacturers to develop appropriate packages
of technology upgrades that combine technologies in ways that work
together and fit with the overall goals of the redesign. It also allows
the manufacturer to fit the process of upgrading emissions control
technology into its multi-year planning process, and it avoids the
large increase in resources and costs that would occur if technology
had to be added outside of the redesign process.
This final rule affects five years of vehicle production, model
years 2012-2016. Given the now-typical five year redesign cycle, nearly
all of a manufacturer's vehicles will be redesigned over this period.
However, this assumes that a manufacturer has sufficient lead time to
redesign the first model year affected by this final rule with the
requirements of this final rule in mind. In fact, the lead time
available for the start of model year 2012 (January 2011) is relatively
short, less than a year. The time between this final rule and the start
of 2013 model year (January 2012) production is under two years. At the
same time, manufacturer product plans indicate that they are planning
on introducing many of the technologies EPA projects could be used to
show compliance with the final CO2 standards in both 2012
and 2013. In order to account for the relatively short lead time
available prior to the 2012 and 2013 model years, albeit mitigated by
their existing plans, EPA has factored this reality into how the
availability is modeled for much of the technology being considered for
model years 2012-2016 as a whole. If the technology to control
greenhouse gas emissions is efficiently folded into this redesign
process, then EPA projects that 85 percent of each manufacturer's sales
will be able to be redesigned with many of the CO2 emission
reducing technologies by the 2016 model year, and as discussed below,
to reduce emissions of HFCs from the air conditioner.
In determining the level of this first ever GHG emissions standard
under the CAA for light-duty vehicles, EPA uses an approach that
accounts for and builds on this redesign process. This provides the
opportunity for several control technologies to be incorporated into
the vehicle during redesign, achieving significant emissions reductions
from the model at one time. This is in contrast to what would be a much
more costly approach of trying to achieve small increments of
reductions over multiple years by adding technology to the vehicle
piece by piece outside of the redesign process.
As described below, the vast majority of technology required by
this final rule is commercially available and already being employed to
a limited extent across the fleet (although the final rule will
necessitate far wider penetration of these technologies throughout the
fleet). The vast majority of the emission reductions which will result
from this final rule will be produced from the increased use of these
technologies. EPA also believes that this final rule will encourage the
development and limited use of more advanced technologies, such as
PHEVs and EVs, and the final rule is structured to facilitate this
result.
In developing the final standard, EPA built on the technical work
performed by the State of California during its development of its
statewide GHG program. EPA began by evaluating a nationwide CAA
standard for MY 2016 that would require the levels of technology
upgrade, across the country, which California standards would

[[Page 25446]]

require for the subset of vehicles sold in California under Pavley 1.
In essence, EPA developed an assessment of an equivalent national new
vehicle fleet-wide CO2 performance standards for model year
2016 which would result in the new vehicle fleet in the State of
California having CO2 performance equal to the performance
from the California Pavley 1 standards. This assessment is documented
in Chapter 3.1 of the RIA. The results of this assessment predicts that
a national light-duty vehicle fleet which adopts technology that
achieves performance of 250 g/mile CO2 for model year 2016
will result in vehicles sold in California that would achieve the
CO2 performance equivalent to the Pavley 1 standards.
EPA then analyzed a level of 250 g/mi CO2 in 2016 using
the OMEGA model (described in more detail below), and the car and truck
footprint curves' relative stringency discussed in Section II to
determine what technology will be needed to achieve a fleet wide
average of 250 g/mi CO2. As discussed later in this section
we believe this level of technology application to the light-duty
vehicle fleet can be achieved in this time frame, that such standards
will produce significant reductions in GHG emissions, and that the
costs for both the industry and the costs to the consumer are
reasonable. EPA also developed standards for the model years 2012
through 2015 that lead up to the 2016 level.
EPA's independent technical assessment of the technical feasibility
of the final MY 2012-2016 standards is described below. EPA has also
evaluated a set of alternative standards for these model years, one
that is more stringent than the final standards and one that is less
stringent. The technical feasibility of these alternative standards is
discussed at the end of this section.
Evaluating the feasibility of these standards primarily includes
identifying available technologies and assessing their effectiveness,
cost, and impact on relevant aspects of vehicle performance and
utility. The wide number of technologies which are available and likely
to be used in combination requires a more sophisticated assessment of
their combined cost and effectiveness. An important factor is also the
degree that these technologies are already being used in the current
vehicle fleet and thus, unavailable for use to improve energy
efficiency beyond current levels. Finally, the challenge for
manufacturers to design the technology into their products, and the
appropriate lead time needed to employ the technology over the product
line of the industry must be considered.
Applying these technologies efficiently to the wide range of
vehicles produced by various manufacturers is a challenging task. In
order to assist in this task, EPA has developed a computerized model
called the Optimization Model for reducing Emissions of Greenhouse
gases from Automobiles (OMEGA) model. Broadly, the model starts with a
description of the future vehicle fleet, including manufacturer, sales,
base CO2 emissions, footprint and the extent to which
emission control technologies are already employed. For the purpose of
this analysis, over 200 vehicle platforms were used to capture the
important differences in vehicle and engine design and utility of
future vehicle sales of roughly 16 million units in the 2016 timeframe.
The model is then provided with a list of technologies which are
applicable to various types of vehicles, along with their cost and
effectiveness and the percentage of vehicle sales which can receive
each technology during the redesign cycle of interest. The model
combines this information with economic parameters, such as fuel prices
and a discount rate, to project how various manufacturers would apply
the available technology in order to meet various levels of emission
control. The result is a description of which technologies are added to
each vehicle platform, along with the resulting cost. While OMEGA can
apply technologies which reduce CO2 emissions and HFC
refrigerant emissions associated with air conditioner use, this task is
currently handled outside of the OMEGA model. The model can be set to
account for various types of compliance flexibilities, such as FFV
credits.
The remainder of this section describes the technical feasibility
analysis in greater detail. Section III.D.1 describes the development
of our projection of the MY 2012-2016 fleet in the absence of this
final rule. Section III.D.2 describes our estimates of the
effectiveness and cost of the control technologies available for
application in the 2012-2016 timeframe. Section III.D.3 combines these
technologies into packages likely to be applied at the same time by a
manufacturer. In this section, the overall effectiveness of the
technology packages vis-[agrave]-vis their effectiveness when combined
individually is described. Section III.D.4 describes the process which
manufacturers typically use to apply new technology to their vehicles.
Section III.D.5 describes EPA's OMEGA model and its approach to
estimating how manufacturers will add technology to their vehicles in
order to comply with CO2 emission standards. Section III.D.6
presents the results of the OMEGA modeling, namely the level of
technology added to manufacturers' vehicles and its cost. Section
III.D.7 discusses the feasibility of the alternative 4-percent-per-year
and 6-percent-per-year standards. Further detail on all of these issues
can be found in EPA and NHTSA's Joint Technical Support Document as
well as EPA's Regulatory Impact Analysis.
1. How did EPA develop a reference vehicle fleet for evaluating further
CO2 reductions?
In order to calculate the impacts of this final rule, it is
necessary to project the GHG emissions characteristics of the future
vehicle fleet absent this regulation. This is called the ``reference''
fleet. EPA and NHTSA develop this reference fleet using a three step
process. Step one develops a set of detailed vehicle characteristics
and sales for a specific model year (in this case, 2008). This is
called the baseline fleet. Step two adjusts the sales of these vehicles
using projections made by AEO and CSM to account for expected changes
in market conditions. Step three applies fuel saving and emission
control technology to these vehicles to the extent necessary for
manufacturers to comply with the MY 2011 CAFE standards. Thus, the
reference fleet differs from the MY 2008 baseline fleet in both the
level of technology utilized and in terms of the sales of any
particular vehicle.
EPA and NHTSA perform steps one and two in an identical manner. The
development of the characteristics of the baseline 2008 fleet and the
adjustment of sales to match AEO and CSM forecasts is described in
detail in Section II.B above. The two agencies perform step three in a
conceptually identical manner, but each agency utilizes its own vehicle
technology and emission model to project the technology needed to
comply with the 2011 CAFE standards. The agencies use the same two
models to project the technology and cost of the 2012-2016 standards.
Use of the same model for both pre-control and post-control costs
ensures consistency.
The agencies received one comment from the Center for Biological
Diversity that the use of 2008 vehicles in our baseline and reference
fleets inherently includes vehicle models which already have or will be
discontinued by the time this rule takes effect and will be replaced by
more advanced vehicle models. This is true. However, we believe that
the use of 2008 vehicle designs is still the most appropriate

[[Page 25447]]

approach available. First, as discussed in Section II.B above, the
designs of these new vehicles at the level of detail required for
emission and cost modeling are not publically available. Even the
confidential descriptions of these vehicle designs are usually not of
sufficient detail to facilitate the level of technology and emission
modeling performed by both agencies. Second, steps two and three of the
process used to create the reference fleet adjust both the sales and
technology of the 2008 vehicles. Thus, our reference fleet reflects the
extent that completely new vehicles are expected to shift the light
vehicle market in terms of both segment and manufacturer. Also, by
adding technology to facilitate compliance with the 2011 CAFE
standards, we account for the vast majority of ways in which these new
vehicles will differ from their older counterparts.
The agencies also received a comment that some manufacturers have
already announced plans to introduce technology well beyond that
required by the 2011 MY CAFE standards. This commenter indicated that
the agencies' approach over-estimated the technology and cost required
by the proposed standards and resulted in less stringent standards
being proposed than a more realistic reference fleet would have
supported. First, the agencies agree that limiting the application of
additional technology beyond that already on 2008 vehicles to only that
required by the 2011 CAFE standards could under-estimate the use of
such technology absent this rule. However, it is difficult, if not
impossible, to separate future fuel economy improvements made for
marketing purposes from those designed to facilitate compliance with
anticipated CAFE or CO2 emission standards. For example,
EISA was signed over two years ago, which contained specific minimum
limits on light vehicle fuel economy in 2020, while also requiring
ratable improvements in the interim. NHTSA proposed fuel economy
standards for the 2012-2015 model years under the EISA provisions in
April of 2008, although NHTSA finalized only 2011 standards for
passenger vehicles. It is also true that manufacturers can change their
plans based on market conditions and other factors. Thus, announcements
of future plans are not certain. As mentioned above, these plans do not
include specific vehicle characteristics. Thus, in order to avoid
under-estimating the cost associated with this rule, the agencies have
limited the fuel economy improvements in the reference fleet to those
projected to result from the existing CAFE standards. We disagree with
the commenter that this has caused the standards being promulgated
today to be less stringent than would have been the case had we been
able to confidently predict additional fuel economy and CO2
emission improvements which will occur absent this rule. The inclusion
of such technology in the reference fleet would certainly have reduced
the cost of this final rule, as well as the benefits, but would not
have changed the final level of technology required to meet the final
standards. Also, we believe that the same impacts would apply to our
evaluations of the two alternative sets of standards, the 4% per year
and 6% per year standards. We are confident that the vast majority of
manufacturers would not comply with the least stringent of these
standards (the 4% per year standards) in the absence of this rule.
Thus, changes to the reference fleet would not have affected the
differences in technology, cost or benefits between the final standards
and the two alternatives. As described below, our rejection of the two
alternatives in favor of the final standards is based primarily on the
relative technology, cost and benefits associated with the three sets
of standards than the absolute cost or benefit relative to the
reference fleet. Thus, we do not agree with the commenter that our
choice of reference fleet adversely impacted the development of the
final standards being promulgated today.
The addition of technology to the baseline fleet so that it
complies with the MY 2011 CAFE standards is described later in Section
III.D.4, as this uses the same methodology used to project compliance
with the final CO2 emission standards. In summary, the
reference fleet represents vehicle characteristics and sales in the
2012 and later model years absent this final rule. Technology is then
added to these vehicles in order to reduce CO2 emissions to
achieve compliance with the final CO2 standards. As noted
above, EPA did not factor in any changes to vehicle utility or
characteristics, or sales in projecting manufacturers' compliance with
this final rule.
After the reference fleet is created, the next step aggregates
vehicle sales by a combination of manufacturer, vehicle platform, and
engine design. As discussed in Section III.D.4 below, manufacturers
implement major design changes at vehicle redesign and tend to
implement these changes across a vehicle platform. Because the cost of
modifying the engine depends on the valve train design (such as SOHC,
DOHC, etc.), the number of cylinders and in some cases head design, the
vehicle sales are broken down beyond the platform level to reflect
relevant engine differences. The vehicle groupings are shown in Table
III.D.1-1. These groupings are the same as those used in the NPRM.

Table III.D.1-1--Vehicle Groupings a
----------------------------------------------------------------------------------------------------------------
Vehicle description Vehicle type Vehicle description Vehicle type
----------------------------------------------------------------------------------------------------------------
Large SUV (Car) V8+ OHV....................... 13 Subcompact Auto I4.............. 1
Large SUV (Car) V6 4v......................... 16 Large Pickup V8+ DOHC........... 19
Large SUV (Car) V6 OHV........................ 12 Large Pickup V8+ SOHC 3v........ 14
Large SUV (Car) V6 2v SOHC.................... 9 Large Pickup V8+ OHV............ 13
Large SUV (Car) I4 and I5..................... 7 Large Pickup V8+ SOHC........... 10
Midsize SUV (Car) V6 2v SOHC.................. 8 Large Pickup V6 DOHC............ 18
Midsize SUV (Car) V6 S/DOHC 4v................ 5 Large Pickup V6 OHV............. 12
Midsize SUV (Car) I4.......................... 7 Large Pickup V6 SOHC 2v......... 11
Small SUV (Car) V6 OHV........................ 12 Large Pickup I4 S/DOHC.......... 7
Small SUV (Car) V6 S/DOHC..................... 4 Small Pickup V6 OHV............. 12
Small SUV (Car) I4............................ 3 Small Pickup V6 2v SOHC......... 8
Large Auto V8+ OHV............................ 13 Small Pickup I4................. 7
Large Auto V8+ SOHC........................... 10 Large SUV V8+ DOHC.............. 17
Large Auto V8+ DOHC, 4v SOHC.................. 6 Large SUV V8+ SOHC 3v........... 14
Large Auto V6 OHV............................. 12 Large SUV V8+ OHV............... 13
Large Auto V6 SOHC 2/3v....................... 5 Large SUV V8+ SOHC.............. 10
Midsize Auto V8+ OHV.......................... 13 Large SUV V6 S/DOHC 4v.......... 16

[[Page 25448]]

Midsize Auto V8+ SOHC......................... 10 Large SUV V6 OHV................ 12
Midsize Auto V7+ DOHC, 4v SOHC................ 6 Large SUV V6 SOHC 2v............ 9
Midsize Auto V6 OHV........................... 12 Large SUV I4.................... 7
Midsize Auto V6 2v SOHC....................... 8 Midsize SUV V6 OHV.............. 12
Midsize Auto V6 S/DOHC 4v..................... 5 Midsize SUV V6 2v SOHC.......... 8
Midsize Auto I4............................... 3 Midsize SUV V6 S/DOHC 4v........ 5
Compact Auto V7+ S/DOHC....................... 6 Midsize SUV I4 S/DOHC........... 7
Compact Auto V6 OHV........................... 12 Small SUV V6 OHV................ 12
Compact Auto V6 S/DOHC 4v..................... 4 Minivan V6 S/DOHC............... 16
Compact Auto I5............................... 7 Minivan V6 OHV.................. 12
Compact Auto I4............................... 2 Minivan I4...................... 7
Subcompact Auto V8+ OHV....................... 13 Cargo Van V8+ OHV............... 13
Subcompact Auto V8+ S/DOHC.................... 6 Cargo Van V8+ SOHC.............. 10
Subcompact Auto V6 2v SOHC.................... 8 Cargo Van V6 OHV................ 12
Subcompact Auto I5/V6 S/DOHC 4v............... 4
----------------------------------------------------------------------------------------------------------------
\a\ I4 = 4 cylinder engine, I5 = 5 cylinder engine, V6, V7, and V8 = 6, 7, and 8 cylinder engines, respectively,
DOHC = Double overhead cam, SOHC = Single overhead cam, OHV = Overhead valve, v = number of valves per
cylinder, ``/'' = and, ``+'' = or larger.

As mentioned above, the second factor which needs to be considered
in developing a reference fleet against which to evaluate the impacts
of this final rule is the impact of the 2011 MY CAFE standards. Since
the vehicles which comprise the above reference fleet are those sold in
the 2008 MY, when coupled with our sales projections, they do not
necessarily meet the 2011 MY CAFE standards.
The levels of the 2011 MY CAFE standards are straightforward to
apply to future sales fleets, as is the potential fine-paying
flexibility afforded by the CAFE program (i.e., $55 per mpg of
shortfall). However, projecting some of the compliance flexibilities
afforded by EISA and the CAFE program are less clear. Two of these
compliance flexibilities are relevant to EPA's analysis: (1) The credit
for FFVs, and (2) the limit on the transferring of credits between car
and truck fleets. The FFV credit is limited to 1.2 mpg in 2011 and EISA
gradually reduces this credit, to 1.0 mpg in 2015 and eventually to
zero in 2020. In contrast, the limit on car-truck transfer is limited
to 1.0 mpg in 2011, and EISA increases this to 1.5 mpg beginning in
2015 and then to 2.0 mpg beginning in 2020. The question here is
whether to hold the 2011 MY CAFE provisions constant in the future or
incorporate the changes in the FFV credit and car-truck credit trading
limits contained in EISA.
As was done for the NPRM, EPA has decided to hold the 2011 MY
limits on FFV credit and car-truck credit trading constant in
projecting the fuel economy and CO2 emission levels of
vehicles in our reference case. This approach treats the changes in the
FFV credit and car-truck credit trading provisions consistently with
the other EISA-mandated changes in the CAFE standards themselves. All
EISA provisions relevant to 2011 MY vehicles are reflected in our
reference case fleet, while all post-2011 MY provisions are not.
Practically, relative to the alternative, this increases both the cost
and benefit of the final standards. In our analysis of this final rule,
any quantified benefits from the presence of FFVs in the fleet are not
considered. Thus, the only impact of the FFV credit is to reduce onroad
fuel economy. By assuming that the FFV credit stays at 1.2 mpg in the
future absent this rule, the assumed level of onroad fuel economy that
would occur absent this final rule is reduced. As this final rule
eliminates the FFV credit (for purposes of CO2 emission
compliance) starting in 2016, the net result is to increase the
projected level of fuel savings from our final standards. Similarly,
the higher level of FFV credit reduces projected compliance cost for
manufacturers to meet the 2011 MY standards in our reference case. This
increases the projected cost of meeting the final 2012 and later
standards.
As just implied, EPA needs to project the technology (and resultant
costs) required for the 2008 MY vehicles to comply with the 2011 MY
CAFE standards in those cases where they do not automatically do so.
The technology and costs are projected using the same methodology that
projects compliance with the final 2012 and later CO2
standards. The description of this process is described in the
following four sections and is essentially the same process used for
the NPRM.
A more detailed description of the methodology used to develop
these sales projections can be found in the Joint TSD. Detailed sales
projections by model year and manufacturer can also be found in the
TSD.
2. What are the effectiveness and costs of CO2-reducing
technologies?
EPA and NHTSA worked together to jointly develop information on the
effectiveness and cost of the CO2-reducing technologies, and
fuel economy-improving technologies, other than A/C related control
technologies. This joint work is reflected in Chapter 3 of the Joint
TSD and in Section II of this preamble. A summary of the effectiveness
and cost of A/C related technology is contained here. For more detailed
information on the effectiveness and cost of A/C related technology,
please refer to Section III.C of this preamble and Chapter 2 of EPA's
RIA.
A/C improvements are an integral part of EPA's technology analysis
and have been included in this section along with the other technology
options. While discussed in Section III.C as a credit opportunity, air
conditioning-related improvements are included in Table III.D.2-1.
because A/C improvements are a very cost-effective technology at
reducing CO2 (or CO2-equivalent) emissions. EPA
expects most manufacturers will choose to use AC improvement credit
opportunities as a strategy for meeting compliance with the
CO2 standards. Note that the costs shown in Table III.D.2-1
do not include maintenance savings that would be expected from the new
AC systems. Further, EPA does not include AC-related maintenance
savings in our cost and benefit analysis presented in Section III.H.
EPA discusses the likely maintenance savings in Chapter 2 of the RIA,
though these savings are not included in our final cost estimates for
the final rule. The EPA approximates that the level of the credits
earned will increase from 2012 to 2016 as more vehicles in the fleet
are redesigned. The

[[Page 25449]]

penetrations and average levels of credit are summarized in Table
III.D.2-2, though the derivation of these numbers (and the breakdown of
car vs. truck credits) is described in the RIA. As demonstrated in the
IMAC study (and described in Section III.C as well as the RIA), these
levels are feasible and achievable with technologies that are available
and cost-effective today.
These improvements are categorized as either leakage reduction,
including use of alternative refrigerants, or system efficiency
improvements. Unlike the majority of the technologies described in this
section, A/C improvements will not be demonstrated in the test cycles
used to quantify CO2 reductions in this final rule. As
described earlier, for this analysis A/C-related CO2
reductions are handled outside of OMEGA model and therefore their
CO2 reduction potential is expressed in grams per mile
rather than a percentage used by the OMEGA model. See Section III.C.1
for the method by which potential reductions are calculated or
measured. Further discussion of the technological basis for these
improvements is included in Chapter 2 of the RIA.

Table III.D.2-1--Total CO2 Reduction Potential and 2016 Cost for A/C
Related Technologies for all Vehicle Classes
[Costs in 2007 dollars]
------------------------------------------------------------------------
CO2 reduction Incremental
potential compliance costs
------------------------------------------------------------------------
A/C refrigerant leakage 7.5 g/mi \249\....... $17
reduction.
A/C efficiency improvements... 5.7 g/mi............. 53
------------------------------------------------------------------------

Table III.D.2-2--A/C Related Technology Penetration and Credit Levels Expected To Be Earned
----------------------------------------------------------------------------------------------------------------
Technology Average credit over entire fleet
penetration --------------------------------------------------------
(percent) Car Truck Fleet average
----------------------------------------------------------------------------------------------------------------
2012................................ \250\ 28 3.4 3.8 3.5
2013................................ 40 4.8 5.4 5.0
2014................................ 60 7.2 8.1 7.5
2015................................ 80 9.6 10.8 10.0
2016................................ 85 10.2 11.5 10.6
----------------------------------------------------------------------------------------------------------------

3. How can technologies be combined into ``packages'' and what is the
cost and effectiveness of packages?
Individual technologies can be used by manufacturers to achieve
incremental CO2 reductions. However, as mentioned in Section
III.D.1, EPA believes that manufacturers are more likely to bundle
technologies into ``packages'' to capture synergistic aspects and
reflect progressively larger CO2 reductions with additions
or changes to any given package. In addition, manufacturers typically
apply new technologies in packages during model redesigns that occur
approximately once every five years, rather than adding new
technologies one at a time on an annual or biennial basis. This way,
manufacturers can more efficiently make use of their redesign resources
and more effectively plan for changes necessary to meet future
standards.
---------------------------------------------------------------------------

\249\ This represents 50% improvement in leakage and thus 50% of
the A/C leakage impact potential compared to a maximum of 15 g/mi
credit that can be achieved through the incorporation of a low very
GWP refrigerant.
\250\ We assume slightly higher A/C penetration in 2012 than was
assumed in the proposal to correct for rounding that occurred in the
curve setting process.
---------------------------------------------------------------------------

Therefore, as explained at proposal, the approach taken here is to
group technologies into packages of increasing cost and effectiveness.
EPA determined that 19 different vehicle types provided adequate
representation to accurately model the entire fleet. This was the
result of analyzing the existing light duty fleet with respect to
vehicle size and powertrain configurations. All vehicles, including
cars and trucks, were first distributed based on their relative size,
starting from compact cars and working upward to large trucks. Next,
each vehicle was evaluated for powertrain, specifically the engine
size, I4, V6, and V8, and finally by the number of valves per cylinder.
Note that each of these 19 vehicle types was mapped into one of the
five classes of vehicles mentioned in Section III.D.2. While the five
classes provide adequate representation for the cost basis associated
with most technology application, they do not adequately account for
all existing vehicle attributes such as base vehicle powertrain
configuration and mass reduction. As an example, costs and
effectiveness estimates for engine friction reduction for the small car
class were used to represent cost and effectiveness for three vehicle
types: Subcompact cars, compact cars, and small multi-purpose vehicles
(MPV) equipped with a 4-cylinder engine, however the mass reduction
associated for each of these vehicle types was based on the vehicle
type sales-weighted average. In another example, a vehicle type for V8
single overhead cam 3-valve engines was created to properly account for
the incremental cost in moving to a dual overhead cam 4-valve
configuration. Note also that these 19 vehicle types span the range of
vehicle footprint (smaller footprints for smaller vehicles and larger
footprints for larger vehicles) which serve as the basis for the
standards being promulgated today. A complete list of vehicles and
their associated vehicle types is shown above in Table III.D.1-1.
Within each of the 19 vehicle types, multiple technology packages
were created in increasing technology content resulting in increasing
effectiveness. Important to note that the effort in creating the
packages attempted to maintain a constant utility for each package as
compared to the baseline package. As such, each package is meant to
provide equivalent driver-perceived performance to the baseline
package. The initial packages represent what a manufacturer will most
likely implement on all vehicles, including low rolling resistance
tires, low friction lubricants, engine friction reduction, aggressive
shift logic, early torque converter lock-up, improved electrical

[[Page 25450]]

accessories, and low drag brakes.\251\ Subsequent packages include
advanced gasoline engine and transmission technologies such as turbo/
downsizing, GDI, and dual-clutch transmission. The most technologically
advanced packages within a segment included HEV, PHEV and EV designs.
The end result is a list of several packages for each of 19 different
vehicle types from which a manufacturer could choose in order to modify
its fleet such that compliance could be achieved.
---------------------------------------------------------------------------

\251\ When making reference to low friction lubricants, the
technology being referred to is the engine changes and possible
durability testing that would be done to accommodate the low
friction lubricants, not the lubricants themselves.
---------------------------------------------------------------------------

Before using these technology packages as inputs to the OMEGA
model, EPA calculated the cost and effectiveness for the package. The
first step was to apply the scaling class for each technology package
and vehicle type combination. The scaling class establishes the cost
and effectiveness for each technology with respect to the vehicle size
or type. The Large Car class was provided as an example in Section
III.D.2. Additional classes include Small Car, Minivan, Small Truck,
and Large Truck and each of the 19 vehicle types was mapped into one of
those five classes. In the next step, the cost for a particular
technology package was determined as the sum of the costs of the
applied technologies. The final step, determination of effectiveness,
requires greater care due to the synergistic effects mentioned in
Section III.D.2. This step is described immediately below.
Usually, the benefits of the engine and transmission technologies
can be combined multiplicatively. For example, if an engine technology
reduces CO2 emissions by five percent and a transmission
technology reduces CO2 emissions by four percent, the
benefit of applying both technologies is 8.8 percent (100%-(100%-4%) *
(100%-5%)). In some cases, however, the benefit of the transmission-
related technologies overlaps with many of the engine technologies.
This occurs because the primary goal of most of the transmission
technologies is to shift operation of the engine to more efficient
locations on the engine map. This is accomplished by incorporating more
ratio selections and a wider ratio span into the transmissions. Some of
the engine technologies have the same goal, such as cylinder
deactivation, advanced valvetrains, and turbocharging. In order to
account for this overlap and avoid over-estimating emissions reduction
effectiveness, EPA has developed a set of adjustment factors associated
with specific pairs of engine and transmission technologies.
The various transmission technologies are generally mutually
exclusive. As such, the effectiveness of each transmission technology
generally supersedes each other. For example, the 9.5-14.5 percent
reduction in CO2 emissions associated with the automated
manual transmission includes the 4.5-6.5 percent benefit of a 6-speed
automatic transmission. Exceptions are aggressive shift logic and early
torque converter lock-up that can be applied to vehicles with several
types of automatic transmissions.
EPA has chosen to use an engineering approach known as the lumped-
parameter technique to determine these adjustment factors. The results
from this approach were then applied directly to the vehicle packages.
The lumped-parameter technique is well documented in the literature,
and the specific approach developed by EPA is detailed in Chapter 1 of
the RIA.
Table III.D.3-1 presents several examples of the reduction in the
effectiveness of technology pairs. A complete list and detailed
discussion of these synergies is presented in Chapter 3 of the Joint
TSD.

Table III.D.3-1--Reduction in Effectiveness for Selected Technology
Pairs
------------------------------------------------------------------------
Reduction in
Transmission combined
Engine technology technology effectiveness
(percent)
------------------------------------------------------------------------
Intake cam phasing............ 5 speed automatic.... 0.5
Coupled cam phasing........... 5 speed automatic.... 0.5
Coupled cam phasing........... Aggressive shift 0.5
logic.
Cylinder deactivation......... 5 speed automatic.... 1.0
Cylinder deactivation......... Aggressive shift 0.5
logic.
------------------------------------------------------------------------

Table III.D.3-2 presents several examples of the CO2-
reducing technology vehicle packages used in the OMEGA model for the
large car class. Similar packages were generated for each of the 19
vehicle types and the costs and effectiveness estimates for each of
those packages are discussed in detail in Chapter 3 of the Joint TSD.

Table III.D.3-2--CO2 Reducing Technology Vehicle Packages for a Large Car Effectiveness and Costs in 2016
[Costs in 2007 dollars]
----------------------------------------------------------------------------------------------------------------
Transmission
Engine technology technology Additional technology CO2 reduction Package cost
----------------------------------------------------------------------------------------------------------------
3.3L V6........................... 4 speed automatic.... None................. Baseline
-----------------------------------------------------------------------------
3.0L V6 + GDI + CCP............... 6 speed automatic.... 3% Mass Reduction.... 17.9% $985
3.0L V6 + GDI + CCP + Deac........ 6 speed automatic.... 5% Mass Reduction.... 20.6% 1,238
2.2L I4 + GDI + Turbo + DCP....... 6 speed DCT.......... 10% Mass Reduction 34.3% 1,903
Start-Stop.
----------------------------------------------------------------------------------------------------------------

[[Page 25451]]

4. Manufacturer's Application of Technology
Vehicle manufacturers often introduce major product changes
together, as a package. In this manner the manufacturers can optimize
their available resources, including engineering, development,
manufacturing and marketing activities to create a product with
multiple new features. In addition, manufacturers recognize that a
vehicle will need to remain competitive over its intended life, meet
future regulatory requirements, and contribute to a manufacturer's CAFE
requirements. Furthermore, automotive manufacturers are largely focused
on creating vehicle platforms to limit the development of entirely new
vehicles and to realize economies of scale with regard to variable
cost. In very limited cases, manufacturers may implement an individual
technology outside of a vehicle's redesign cycle.\252\ In following
with these industry practices, EPA has created set of vehicle
technology packages that represent the entire light duty fleet.
---------------------------------------------------------------------------

\252\ The Center for Biological Diversity submitted comments
disputing this distinction as well as the need for lead time. These
comments are addressed in Section III.D.7.
---------------------------------------------------------------------------

In evaluating needed lead time, EPA has historically authorized
manufacturers of new vehicles or nonroad equipment to phase in
available emission control technology over a number of years. Examples
of this are EPA's Tier 2 program for cars and light trucks and its 2007
and later PM and NOX emission standards for heavy-duty
vehicles. In both of these rules, the major modifications expected from
the rules were the addition of exhaust aftertreatment control
technologies. Some changes to the engine were expected as well, but
these were not expected to affect engine size, packaging or
performance. The CO2 reduction technologies described above
potentially involve much more significant changes to car and light
truck designs. Many of the engine technologies involve changes to the
engine block and heads. The transmission technologies could change the
size and shape of the transmission and thus, packaging. Improvements to
aerodynamic drag could involve body design and therefore, the dies used
to produce body panels. Changes of this sort potentially involve new
capital investment and the obsolescence of existing investment.
At the same time, vehicle designs are not static, but change in
major ways periodically. The manufacturers' product plans indicate that
vehicles are usually redesigned every 5 years on average.\253\ Vehicles
also tend to receive a more modest ``refresh'' between major redesigns,
as discussed above. Because manufacturers are already changing their
tooling, equipment and designs at these times, further changes to
vehicle design at these times involve a minimum of stranded capital
equipment. Thus, the timing of any major technological changes is
projected to coincide with changes that manufacturers are already
making to their vehicles. This approach effectively avoids the need to
quantify any costs associated with discarding equipment, tooling,
emission and safety certification, etc. when CO2-reducing
equipment is incorporated into a vehicle.
---------------------------------------------------------------------------

\253\ See discussion in Section III.D.7 with references.
---------------------------------------------------------------------------

This final rule affects five years of vehicle production, model
years 2012-2016. Given the now-typical five year redesign cycle, nearly
all of a manufacturer's vehicles will be redesigned over this period.
However, this assumes that a manufacturer has sufficient lead time to
redesign the first model year affected by this final rule with the
requirements of this final rule in mind. In fact, the lead time
available for model year 2012 is relatively short. The time between a
likely final rule and the start of 2013 model year production is likely
to be just over two years. At the same time, the manufacturer product
plans indicate that they are planning on introducing many of the
technologies projected to be required by this final rule in both 2012
and 2013. In order to account for the relatively short lead time
available prior to the 2012 and 2013 model years, albeit mitigated by
their existing plans, EPA projects that only 85 percent of each
manufacturer's sales will be able to be redesigned with major
CO2 emission-reducing technologies by the 2016 model year.
Less intrusive technologies can be introduced into essentially all of a
manufacturer's sales. This resulted in three levels of technology
penetration caps, by manufacturer. Common technologies (e.g., low
friction lubes, aerodynamic improvements) had a penetration cap of
100%. More advanced powertrain technologies (e.g., stoichiometric GDI,
turbocharging) had a penetration cap of 85%. The most advanced
technologies considered in this analysis (e.g., diesel engines,\254\ as
well as IMA, powersplit and 2-mode hybrids) had a 15% penetration cap.
---------------------------------------------------------------------------

\254\ While diesel engines are a mature technology and not
``advanced'', the aftertreatment systems necessary for them in the
U.S. market are advanced.
---------------------------------------------------------------------------

This is the same approach as was taken in the NPRM. EPA received
several comments commending it on its approach to establishing
technical feasibility via its use of the OMEGA model. The only adverse
comment received regarding the application of technology was from the
Center for Biological Diversity (CBD), which criticized EPA's use of
the 5-year redesign cycle. CBD argued that manufacturers occasionally
redesign vehicles sooner than 5 years and that EPA did not quantify the
cost of shortening the redesign cycle to less than 5 years and compare
this cost to the increased benefit of reduced CO2 emissions.
CBD also noted that manufacturers have been recently dropping vehicle
lines and entire divisions with very little leadtime, indicating their
ability to change product plans much quicker than projected above.
EPA did not explicitly evaluate the cost of reducing the average
redesign cycle to less than 5 years for two reasons. One, in the past,
manufacturers have usually shortened the redesign cycle to address
serious problems with the current design, usually lower than
anticipated sales. However, the amortized cost of the capital necessary
to produce a new vehicle design will increase by 23%, from one-fifth of
the capital cost to one-fourth (and assuming a 3% discount rate). This
would be on top of the cost of the emission control equipment itself.
The only benefit of this increase in societal cost will be earlier
CO2 emission reductions (and the other benefits associated
with CO2 emission control). The capital costs associated
with vehicle redesign go beyond CO2 emission control and
potentially involve every aspect of the vehicle and can represent
thousands of dollars. We believe that it would be an inefficient use of
societal resources to incur such costs when they can be obtained much
more cost effectively just one year later.
Two, the examples of manufacturers dropping vehicle lines and
divisions with very short lead time is not relevant to the redesign of
vehicles. There is no relationship between a manufacturer's ability to
stop selling a vehicle model or to close a vehicle division and a
manufacturer's ability to redesign a vehicle. A company could decide to
stop selling all of its products within a few weeks--but it would still
take a firm approximately 5 years to introduce a major new vehicle
line. It is relatively easy to stop the manufacture of a particular
product (though this too can

[[Page 25452]]

incur some cost--such as plant wind-down costs, employee layoff or
relocation costs, and dealership related costs). It is much more
difficult to perform the required engineering design and development,
design, purchase, and install the necessary capital equipment and
tooling for components and vehicle manufacturing and develop all the
processes associated with the application of a new technology. Further
discussion of the CBD comments can be found in III.D.7 below.
5. How is EPA projecting that a manufacturer decides between options to
improve CO2 performance to meet a fleet average standard?
EPA is generally taking the same approach to projecting the
application of technology to vehicles as it did for the NPRM. With the
exception of two comments, all commenters agreed with the modeling
approach taken in the NPRM. One of these two comments is addressed is
Section III.D.1 above, while the other is addressed in Section III.D.3.
above.
There are many ways for a manufacturer to reduce CO2-
emissions from its vehicles. A manufacturer can choose from a myriad of
CO2 reducing technologies and can apply one or more of these
technologies to some or all of its vehicles. Thus, for a variety of
levels of CO2 emission control, there are an almost infinite
number of technology combinations which produce the desired
CO2 reduction. As noted earlier, EPA developed a new vehicle
model, the OMEGA model in order to make a reasonable estimate of how
manufacturers will add technologies to vehicles in order to meet a
fleet-wide CO2 emissions level. EPA has described OMEGA's
specific methodologies and algorithms in a memo to the docket for this
rulemaking (Docket EPA-HQ-OAR-2009-0472).
The OMEGA model utilizes four basic sets of input data. The first
is a description of the vehicle fleet. The key pieces of data required
for each vehicle are its manufacturer, CO2 emission level,
fuel type, projected sales and footprint. The model also requires that
each vehicle be assigned to one of the 19 vehicle types, which tells
the model which set of technologies can be applied to that vehicle.
(For a description of how the 19 vehicle types were created, reference
Section III.D.3.) In addition, the degree to which each vehicle already
reflects the effectiveness and cost of each available technology must
also be input. This avoids the situation, for example, where the model
might try to add a basic engine improvement to a current hybrid
vehicle. Except for this type of information, the development of the
required data regarding the reference fleet was described in Section
III.D.1 above and in Chapter 1 of the Joint TSD.
The second type of input data used by the model is a description of
the technologies available to manufacturers, primarily their cost and
effectiveness. Note that the five vehicle classes are not explicitly
used by the model, rather the costs and effectiveness associated with
each vehicle package is based on the associated class. This information
was described in Sections III.D.2 and III.D.3 above as well as Chapter
3 of the Joint TSD. In all cases, the order of the technologies or
technology packages for a particular vehicle type is determined by the
model user prior to running the model. Several criteria can be used to
develop a reasonable ordering of technologies or packages. These are
described in the Joint TSD.
The third type of input data describes vehicle operational data,
such as annual scrap rates and mileage accumulation rates, and economic
data, such as fuel prices and discount rates. These estimates are
described in Section II.F above, Section III.H below and Chapter 4 of
the Joint TSD.
The fourth type of data describes the CO2 emission
standards being modeled. These include the CO2 emission
equivalents of the 2011 MY CAFE standards and the final CO2
standards for 2016. As described in more detail below, the application
of A/C technology is evaluated in a separate analysis from those
technologies which impact CO2 emissions over the 2-cycle
test procedure. Thus, for the percent of vehicles that are projected to
achieve A/C related reductions, the CO2 credit associated
with the projected use of improved A/C systems is used to adjust the
final CO2 standard which will be applicable to each
manufacturer to develop a target for CO2 emissions over the
2-cycle test which is assessed in our OMEGA modeling.
As mentioned above for the market data input file utilized by
OMEGA, which characterizes the vehicle fleet, our modeling must and
does account for the fact that many 2008 MY vehicles are already
equipped with one or more of the technologies discussed in Section
III.D.2 above. Because of the choice to apply technologies in packages,
and 2008 vehicles are equipped with individual technologies in a wide
variety of combinations, accounting for the presence of specific
technologies in terms of their proportion of package cost and
CO2 effectiveness requires careful, detailed analysis. The
first step in this analysis is to develop a list of individual
technologies which are either contained in each technology package, or
would supplant the addition of the relevant portion of each technology
package. An example would be a 2008 MY vehicle equipped with variable
valve timing and a 6-speed automatic transmission. The cost and
effectiveness of variable valve timing would be considered to be
already present for any technology packages which included the addition
of variable valve timing or technologies which went beyond this
technology in terms of engine related CO2 control
efficiency. An example of a technology which supplants several
technologies would be a 2008 MY vehicle which was equipped with a
diesel engine. The effectiveness of this technology would be considered
to be present for technology packages which included improvements to a
gasoline engine, since the resultant gasoline engines have a lower
CO2 control efficiency than the diesel engine. However, if
these packages which included improvements also included improvements
unrelated to the engine, like transmission improvements, only the
engine related portion of the package already present on the vehicle
would be considered. The transmission related portion of the package's
cost and effectiveness would be allowed to be applied in order to
comply with future CO2 emission standards.
The second step in this process is to determine the total cost and
CO2 effectiveness of the technologies already present and
relevant to each available package. Determining the total cost usually
simply involves adding up the costs of the individual technologies
present. In order to determine the total effectiveness of the
technologies already present on each vehicle, the lumped parameter
model described above is used. Because the specific technologies
present on each 2008 vehicle are known, the applicable synergies and
dis-synergies can be fully accounted for.
The third step in this process is to divide the total cost and
CO2 effectiveness values determined in step 2 by the total
cost and CO2 effectiveness of the relevant technology
packages. These fractions are capped at a value of 1.0 or less, since a
value of 1.0 causes the OMEGA model to not change either the cost or
CO2 emissions of a vehicle when that technology package is
added.
As described in Section III.D.3 above, technology packages are
applied to groups of vehicles which generally represent a single
vehicle platform and which are equipped with a single engine size
(e.g., compact cars with four cylinder engine produced by Ford). These
grouping are described in Table III.D.1-1. Thus, the fourth step is to

[[Page 25453]]

combine the fractions of the cost and effectiveness of each technology
package already present on the individual 2008 vehicles models for each
vehicle grouping. For cost, percentages of each package already present
are combined using a simple sales-weighting procedure, since the cost
of each package is the same for each vehicle in a grouping. For
effectiveness, the individual percentages are combined by weighting
them by both sales and base CO2 emission level. This
appropriately weights vehicle models with either higher sales or
CO2 emissions within a grouping. Once again, this process
prevents the model from adding technology which is already present on
vehicles, and thus ensures that the model does not double count
technology effectiveness and cost associated with complying with the
2011 MY CAFE standards and the final CO2 standards.
Conceptually, the OMEGA model begins by determining the specific
CO2 emission standard applicable for each manufacturer and
its vehicle class (i.e., car or truck). Since the final rule allows for
averaging across a manufacturer's cars and trucks, the model determines
the CO2 emission standard applicable to each manufacturer's
car and truck sales from the two sets of coefficients describing the
piecewise linear standard functions for cars and trucks in the inputs,
and creates a combined car-truck standard. This combined standard
considers the difference in lifetime VMT of cars and trucks, as
indicated in the final regulations which govern credit trading between
these two vehicle classes. For both the 2011 CAFE and 2016
CO2 standards, these standards are a function of each
manufacturer's sales of cars and trucks and their footprint values.
When evaluating the 2011 MY CAFE standards, the car-truck trading was
limited to 1.2 mpg. When evaluating the final CO2 standards,
the OMEGA model was run only for MY 2016. OMEGA is designed to evaluate
technology addition over a complete redesign cycle and 2016 represents
the final year of a redesign cycle starting with the first year of the
final CO2 standards, 2012. Estimates of the technology and
cost for the interim model years are developed from the model
projections made for 2016. This process is discussed in Chapter 6 of
EPA's RIA to this final rule. When evaluating the 2016 standards using
the OMEGA model, the final CO2 standard which manufacturers
will otherwise have to meet to account for the anticipated level of A/C
credits generated was adjusted. On an industry wide basis, the
projection shows that manufacturers will generate 11 g/mi of A/C credit
in 2016. Thus, the 2016 CO2 target for the fleet evaluated
using OMEGA was 261 g/mi instead of 250 g/mi.
As noted above, EPA estimated separately the cost of the improved
A/C systems required to generate the 11 g/mi credit. This is consistent
with our final A/C credit procedures, which will grant manufacturers A/
C credits based on their total use of improved A/C systems, and not on
the increased use of such systems relative to some base model year
fleet. Some manufacturers may already be using improved A/C technology.
However, this represents a small fraction of current vehicle sales. To
the degree that such systems are already being used, EPA is over-
estimating both the cost and benefit of the addition of improved A/C
technology relative to the true reference fleet to a small degree.
The model then works with one manufacturer at a time to add
technologies until that manufacturer meets its applicable standard. The
OMEGA model can utilize several approaches to determining the order in
which vehicles receive technologies. For this analysis, EPA used a
``manufacturer-based net cost-effectiveness factor'' to rank the
technology packages in the order in which a manufacturer is likely to
apply them. Conceptually, this approach estimates the cost of adding
the technology from the manufacturer's perspective and divides it by
the mass of CO2 the technology will reduce. One component of
the cost of adding a technology is its production cost, as discussed
above. However, it is expected that new vehicle purchasers value
improved fuel economy since it reduces the cost of operating the
vehicle. Typical vehicle purchasers are assumed to value the fuel
savings accrued over the period of time which they will own the
vehicle, which is estimated to be roughly five years. It is also
assumed that consumers discount these savings at the same rate as that
used in the rest of the analysis (3 or 7 percent). Any residual value
of the additional technology which might remain when the vehicle is
sold is not considered. The CO2 emission reduction is the
change in CO2 emissions multiplied by the percentage of
vehicles surviving after each year of use multiplied by the annual
miles travelled by age, again discounted to the year of vehicle
purchase.
Given this definition, the higher priority technologies are those
with the lowest manufacturer-based net cost-effectiveness value
(relatively low technology cost or high fuel savings leads to lower
values). Because the order of technology application is set for each
vehicle, the model uses the manufacturer-based net cost-effectiveness
primarily to decide which vehicle receives the next technology
addition. Initially, technology package 1 is the only one
available to any particular vehicle. However, as soon as a vehicle
receives technology package 1, the model considers the
manufacturer-based net cost-effectiveness of technology package
2 for that vehicle and so on. In general terms, the equation
describing the calculation of manufacturer-based cost effectiveness is
as follows:
[GRAPHIC] [TIFF OMITTED] TR07MY10.018

Where

ManufCostEff = Manufacturer-Based Cost Effectiveness (in dollars per
kilogram CO2),
TechCost = Marked up cost of the technology (dollars),
PP = Payback period, or the number of years of vehicle use over
which consumers value fuel savings when evaluating the value of a
new vehicle at time of purchase,
dFSi = Difference in fuel consumption due to the addition
of technology times fuel price in year i,
dCO2 = Difference in CO2 emissions due to the
addition of technology,
VMTi = product of annual VMT for a vehicle of age i and the
percentage of vehicles of age i still on the road, and
1- Gap = Ratio of onroad fuel economy to two-cycle (FTP/HFET) fuel
economy.

[[Page 25454]]

The OMEGA model does not currently allow for the VMT used in
determining the various technology ranking factors to be a function of
the rebound factor. If the user believed that the consideration of
rebound VMT was important, they could increase their estimate of the
payback period to simulate the impact of the rebound VMT.
EPA describes the technology ranking methodology and manufacturer-
based cost effectiveness metric in greater detail in a technical memo
to the Docket for this final rule (Docket EPA-HQ-OAR-2009-0472).
When calculating the fuel savings, the full retail price of fuel,
including taxes is used. While taxes are not generally included when
calculating the cost or benefits of a regulation, the net cost
component of the manufacturer-based net cost-effectiveness equation is
not a measure of the social cost of this final rule, but a measure of
the private cost, (i.e., a measure of the vehicle purchaser's
willingness to pay more for a vehicle with higher fuel efficiency).
Since vehicle operators pay the full price of fuel, including taxes,
they value fuel costs or savings at this level, and the manufacturers
will consider this when choosing among the technology options.
This definition of manufacturer-based net cost-effectiveness
ignores any change in the residual value of the vehicle due to the
additional technology when the vehicle is five years old. As discussed
in Chapter 1 of the RIA, based on historic used car pricing, applicable
sales taxes, and insurance, vehicles are worth roughly 23% of their
original cost after five years, discounted to year of vehicle purchase
at 7% per annum. It is reasonable to estimate that the added technology
to improve CO2 level and fuel economy will retain this same
percentage of value when the vehicle is five years old. However, it is
less clear whether first purchasers, and thus, manufacturers consider
this residual value when ranking technologies and making vehicle
purchases, respectively. For this final rule, this factor was not
included in our determination of manufacturer-based net cost-
effectiveness in the analyses performed in support of this final rule.
The values of manufacturer-based net cost-effectiveness for
specific technologies will vary from vehicle to vehicle, often
substantially. This occurs for three reasons. First, both the cost and
fuel-saving component cost, ownership fuel-savings, and lifetime
CO2 effectiveness of a specific technology all vary by the
type of vehicle or engine to which it is being applied (e.g., small car
versus large truck, or 4-cylinder versus 8-cylinder engine). Second,
the effectiveness of a specific technology often depends on the
presence of other technologies already being used on the vehicle (i.e.,
the dis-synergies). Third, the absolute fuel savings and CO2
reduction of a percentage on incremental reduction in fuel consumption
depends on the CO2 level of the vehicle prior to adding the
technology. Chapter 1 of the RIA of this final rule contains further
detail on the values of manufacturer-based net cost-effectiveness for
the various technology packages.
6. Why are the final CO2 standards feasible?
The finding that the final standards are technically feasible is
based primarily on two factors. One is the level of technology needed
to meet the final standards. The other is the cost of this technology.
The focus is on the final standards for 2016, as this is the most
stringent standard and requires the most extensive use of technology.
With respect to the level of technology required to meet the
standards, EPA established technology penetration caps. As described in
Section III.D.4, EPA used two constraints to limit the model's
application of technology by manufacturer. The first was the
application of common fuel economy enablers such as low rolling
resistance tires and transmission logic changes. These were allowed to
be used on all vehicles and hence had no penetration cap. The second
constraint was applied to most other technologies and limited their
application to 85% with the exception of the most advanced technologies
(e.g., power-split hybrid and 2-mode hybrid) and diesel,\255\ whose
application was limited to 15%.
---------------------------------------------------------------------------

\255\ While diesel engines are not an ``advanced technology''
per se, diesel engines that can meet EPA's light duty Tier 2 Bin 5
NOX standards have advanced (and somewhat costly)
aftertreatment systems on them that make this technology penetration
cap appropriate in addition to their relatively high incremental
costs.
---------------------------------------------------------------------------

EPA used the OMEGA model to project the technology (and resultant
cost) required for manufacturers to meet the current 2011 MY CAFE
standards and the final 2016 MY CO2 emission standards. Both
sets of standards were evaluated using the OMEGA model. The 2011 MY
CAFE standards were applied to cars and trucks separately with the
transfer of credits from one category to the other allowed up to an
increase in fuel economy of 1.0 mpg as allowed under the applicable MY
2011 CAFE regulations. Chrysler, Ford and General Motors are assumed to
utilize FFV credits up to the maximum of 1.2 mpg for both their car and
truck sales. Nissan is assumed to utilize FFV credits up to the maximum
of 1.2 mpg for only their truck sales. The use of any banked credits
from previous model years was not considered. The modification of the
reference fleet to comply with the 2011 CAFE standards through the
application of technology by the OMEGA model is the final step in
creating the final reference fleet. This final reference fleet forms
the basis for comparison for the model year 2016 standards.
Table III.D.6-1 shows the usage level of selected technologies in
the 2008 vehicles coupled with 2016 sales prior to projecting their
compliance with the 2011 MY CAFE standards. These technologies include
converting port fuel-injected gasoline engines to direct injection
(GDI), adding the ability to deactivate certain engine cylinders during
low load operation to overhead cam engines (OHC-DEAC), adding a
turbocharger and downsizing the engine (Turbo), diesel engine
technology, increasing the number of transmission speeds to 6, or
converting automatic transmissions to dual-clutch automated manual
transmissions (Dual-Clutch Trans), adding 42 volt start-stop capability
(Start-Stop), and converting a vehicle to an intermediate or strong
hybrid design. This last category includes three current hybrid
designs: Integrated motor assist (IMA), power-split (PS), 2-mode
hybrids and electric vehicles.\256\
---------------------------------------------------------------------------

\256\ EPA did not project reliance on the use of any plug-in
hybrid or battery electric vehicles when projecting manufacturers'
compliance with the 2016 standards. However, BMW did sell a battery
electric vehicle in the 2008 model year, so these sales are included
in the technology penetration estimates for the reference case and
the final and alternative standards evaluated for 2016.

[[Page 25455]]

Table III.D.6-1--Penetration of Technology in 2008 Vehicles With 2016 Sales: Cars and Trucks
[Percent of sales]
--------------------------------------------------------------------------------------------------------------------------------------------------------
6 Speed Dual clutch
GDI OHC-DEAC Turbo Diesel auto trans trans Start-stop Hybrid
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW............................................. 7.5 0.0 6.1 0.0 86 0.9 0 0.1
Chrysler........................................ 0.0 0.0 0.5 0.1 14 0.0 0 0.0
Daimler......................................... 0.0 0.0 6.5 5.6 76 7.5 0 0.0
Ford............................................ 0.4 0.0 2.2 0.0 29 0.0 0 0.0
General Motors.................................. 3.1 0.0 1.4 0.0 15 0.0 0 0.3
Honda........................................... 1.4 7.1 1.4 0.0 0 0.0 0 2.1
Hyundai......................................... 0.0 0.0 0.0 0.0 3 0.0 0 0.0
Kia............................................. 0.0 0.0 0.0 0.0 0 0.0 0 0.0
Mazda........................................... 13.6 0.0 13.6 0.0 26 0.0 0 0.0
Mitsubishi...................................... 0.0 0.0 0.0 0.0 10 0.0 0 0.0
Nissan.......................................... 0.0 0.0 0.0 0.0 0 0.0 0 0.8
Porsche......................................... 58.6 0.0 14.9 0.0 49 0.0 0 0.0
Subaru.......................................... 0.0 0.0 9.8 0.0 0 0.0 0 0.0
Suzuki.......................................... 0.0 0.0 0.0 0.0 0 0.0 0 0.0
Tata............................................ 0.0 0.0 17.3 0.0 99 0.0 0 0.0
Toyota.......................................... 6.8 0.0 0.0 0.0 21 0.0 0 11.6
Volkswagen...................................... 50.6 0.0 39.5 0.0 69 13.1 0 0.0
Overall......................................... 3.8 0.8 2.6 0.1 19.1 0.5 0.0 2.2
--------------------------------------------------------------------------------------------------------------------------------------------------------

As can be seen, all of these technologies were already being used
on some 2008 MY vehicles, with the exception of direct injection
gasoline engines with either cylinder deactivation or turbocharging and
downsizing. Transmissions with more gearsets were the most prevalent,
with some manufacturers (e.g., BMW, Suzuki) using them on essentially
all of their vehicles. Both Daimler and VW equip many of their vehicles
with automated manual transmissions, while VW makes extensive use of
direct injection gasoline engine technology. Toyota has converted a
significant percentage of its 2008 vehicles to strong hybrid design.
Table III.D.6-2 shows the usage level of the same technologies in
the reference case fleet after projecting their compliance with the
2011 MY CAFE standards. Except for mass reduction, the figures shown
represent the percentages of each manufacturer's sales which are
projected to be equipped with the indicated technology. For mass
reduction, the overall mass reduction projected for that manufacturer's
sales is also shown. The last row in Table III.D.6-2 shows the increase
in projected technology penetration due to compliance with the 2011 MY
CAFE standards. The results of DOT's Volpe modeling were used to
project that all manufacturers would comply with the 2011 MY standards
in 2016 without the need to pay fines, with one exception. This
exception was Porsche in the case of their car fleet. When projecting
Porsche's compliance with the 2011 MY CAFE standard for cars, NHTSA
projected that Porsche would achieve a CO2 emission level of
304.3 g/mi instead of the required 284.8 g/mi level (29.2 mpg instead
of 31.2 mpg), and pay fines in lieu of further control.

Table III.D.6-2--Penetration of Technology Under 2011 MY CAFE Standards in 2016 Sales: Cars and Trucks
[Percent of sales]
--------------------------------------------------------------------------------------------------------------------------------------------------------
6 Speed Dual clutch Mass
GDI OHC-DEAC Turbo auto trans trans Start-stop reduction
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW.......................................................... 44 12 30 53 37 13 2
Chrysler..................................................... 0 0 0 18 0 0 0
Daimler...................................................... 23 22 8 52 34 26 2
Ford......................................................... 0 0 3 27 0 0 0
General Motors............................................... 3 0 1 15 0 0 0
Honda........................................................ 2 6 2 0 0 0 0
Hyundai...................................................... 0 0 0 3 0 0 0
Kia.......................................................... 0 0 0 0 0 0 0
Mazda........................................................ 13 0 13 20 0 0 0
Mitsubishi................................................... 32 0 2 25 35 0 1
Nissan....................................................... 0 0 0 0 0 0 0
Porsche...................................................... 92 0 75 5 55 38 4
Subaru....................................................... 0 0 9 0 0 0 0
Suzuki....................................................... 70 0 0 3 67 67 3
Tata......................................................... 85 54 20 27 73 73 6
Toyota....................................................... 7 0 0 19 0 0 0
Volkswagen................................................... 89 5 81 14 78 18 3
Overall...................................................... 10 2 7 16 7 3 0
Increase over 2008 MY........................................ 6 1 4 -3 6 3 0
--------------------------------------------------------------------------------------------------------------------------------------------------------

[[Page 25456]]

As can be seen, the 2011 MY CAFE standards, when evaluated on an
industry wide basis, require only a modest increase in the use of these
technologies. The projected MY 2016 fraction of automatic transmission
with more gearsets actually decreases slightly due to conversion of
these units to more efficient designs such as automated manual
transmissions and hybrids. However, the impact of the 2011 MY CAFE
standards is much greater on selected manufacturers, particularly BMW,
Daimler, Porsche, Tata (Jaguar/Land Rover) and VW. All of these
manufacturers are projected to increase their use of direct injection
gasoline engine technology, advanced transmission technology, and
start-stop technology. It should be noted that these manufacturers have
traditionally paid fines under the CAFE program. However, with higher
fuel prices and the lower cost mature technology projected to be
available by 2016, these manufacturers would likely find it in their
best interest to improve their fuel economy levels instead of
continuing to pay fines (again with the exception of Porsche cars).
While not shown, no gasoline engines were projected to be converted to
diesel technology and no hybrid vehicles were projected. Most
manufacturers do not require the level of CO2 emission
control associated with either of these technologies. The few
manufacturers that would were projected to choose to pay CAFE fines in
2011 in lieu of adding diesel or hybrid technologies.
This 2008 baseline fleet, modified to meet 2011 standards, becomes
our ``reference'' case. See Section II.B above. This is the fleet
against which the final 2016 standards are compared. Thus, it is also
the fleet that is assumed to exist in the absence of this rule. No air
conditioning improvements are assumed for model year 2011 vehicles. The
average CO2 emission levels of this reference fleet vary
slightly from 2012-2016 due to small changes in the vehicle sales by
market segments and manufacturer. CO2 emissions from cars
range from 282-284 g/mi, while those from trucks range from 382-384 g/
mi. CO2 emissions from the combined fleet range from 316-
320. These estimates are described in greater detail in Section 5.3.2.2
of the EPA RIA.
Conceptually, both EPA and NHTSA perform the same projection in
order to develop their respective reference fleets. However, because
the two agencies use two different models to modify the baseline fleet
to meet the 2011 CAFE standards, the projected technology that could be
added will be slightly different. The differences, however, are
relatively small since most manufacturers only require modest addition
of technology to meet the 2011 CAFE standards.
EPA then used the OMEGA model once again to project the level of
technology needed to meet the final 2016 CO2 emission
standards. Using the results of the OMEGA model, every manufacturer was
projected to be able to meet the final 2016 standards with the
technology described above except for four: BMW, VW, Porsche and Tata
(which is comprised of Jaguar and Land Rover vehicles in the U.S.
fleet). For these manufacturers, the results presented below are those
with the fully allowable application of technology available in EPA's
OMEGA modeling analysis and not for the technology projected to enable
compliance with the final standards. Described below are a number of
potential feasible solutions for how these companies can achieve
compliance. The overall level of technology needed to meet the final
2016 standards is shown in Table III.D.6-3. As discussed above, all
manufacturers are projected to improve the air conditioning systems on
85% of their 2016 sales.\257\
---------------------------------------------------------------------------

\257\ Many of the technologies shown in this table are mutually
exclusive. Thus, 85% penetration might not be possible. For example,
any use of hybrids will reduce the DEAC, Turbo, 6SPD, DCT, and 42V
S-S technologies. Additionally, not every technology is available to
be used on every vehicle type.

Table III.D.6-3--Final Penetration of Technology for 2016 CO2 Standards: Cars and Trucks
[Percent of sales]
--------------------------------------------------------------------------------------------------------------------------------------------------------
6 Speed Dual clutch Mass
GDI OHC-DEAC Turbo Diesel auto trans trans Start-stop Hybrid Reduction
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW................................ 80 21 61 6 13 63 65 14 5
Chrysler........................... 79 13 17 0 31 52 54 0 6
Daimler............................ 76 30 53 5 12 72 67 14 5
Ford............................... 84 21 19 0 27 60 61 0 6
General Motors..................... 67 25 14 0 8 61 61 0 6
Honda.............................. 43 6 2 0 0 49 18 2 3
Hyundai............................ 59 0 1 0 8 52 32 0 3
Kia................................ 33 0 1 0 0 52 4 0 2
Mazda.............................. 60 0 14 1 17 47 41 0 4
Mitsubishi......................... 74 0 33 0 14 74 74 0 6
Nissan............................. 66 7 11 0 2 62 58 1 5
Porsche............................ 83 15 62 8 5 45 62 15 4
Subaru............................. 60 0 9 0 0 58 44 0 3
Suzuki............................. 77 0 0 0 10 67 67 0 4
Tata............................... 85 55 27 0 14 70 70 15 5
Toyota............................. 26 7 3 0 13 40 7 12 2
Volkswagen......................... 82 18 71 11 10 68 60 15 4
Overall............................ 60 13 15 1 12 55 42 4 4
Increase over 2011 CAFE............ 49 11 9 1 -4 48 39 2 4
--------------------------------------------------------------------------------------------------------------------------------------------------------

[[Page 25457]]

Table III.D.6-4 shows the 2016 standards, as well as the achieved
CO2 emission levels for the five manufacturers which are not
able to meet these standards under the premises of our modeling. It
should be noted that the two sets of combined emission levels shown in
Table III.D.6-4 are based on sales weighting car and truck emission
levels.

Table III.D.6-4--Emissions of Manufacturers Unable to Meet Final 2016 Standards (g/mi CO2)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Achieved emissions 2016 Standards Shortfall
Manufacturer -------------------------------------------------------------------------------------------------
Car Truck Combined Car Truck Combined Combined
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW................................................... 236.3 278.7 248.5 228.4 282.5 243.9 4.6
Tata.................................................. 258.6 323.6 284.2 249.9 272.5 258.8 25.4
Daimler............................................... 246.3 297.8 262.6 238.3 294.3 256.1 6.5
Porsche............................................... 244.1 332.0 273.4 206.1 286.9 233.0 40.4
Volkswagen............................................ 223.5 326.6 241.6 218.6 292.7 231.6 10.0
--------------------------------------------------------------------------------------------------------------------------------------------------------

As can be seen, BMW and Daimler have the smallest shortfalls, 5-6
g/mi, while Porsche has the largest, 40 g/mi.
On an industry average basis, the technology penetrations are very
similar to those projected in the proposal. There is a slight shift
from the use of cylinder deactivation to the two advanced transmission
technologies. This is due to the fact that the estimated costs for
these three technologies have been updated, and thus, their relative
cost effectiveness when applied to specific vehicles have also shifted.
The reader is referred to Section II.E of this preamble as well as
Chapter 3 of the Joint TSD for a detailed description of the cost
estimates supporting this final rule and to the RIA for a description
of the selection of technology packages for specific vehicle types. The
other technologies shown in Table III.D.6-4 changed by 2 percent or
less between the proposal and this final rule.
As can be seen, the overall average reduction in vehicle weight is
projected to be 4 percent. This reduction varies across the two vehicle
classes and vehicle base weight. For cars below 2,950 pounds curb
weight, the average reduction is 2.8 percent (75 pounds), while the
average was 4.3 percent (153 pounds) for cars above 2,950 curb weight.
For trucks below 3,850 pounds curb weight, the average reduction is 4.7
percent (163 pounds), while it was 5.1 percent (240 pounds) for trucks
above 3,850 curb weight. Splitting trucks at a higher weight, for
trucks below 5,000 pounds curb weight, the average reduction is 4.4
percent (186 pounds), while it was 7.0 percent (376 pounds) for trucks
above 5,000 curb weight.
The levels of requisite technologies differ significantly across
the various manufacturers. Therefore, several analyses were performed
to ascertain the cause. Because the baseline case fleet consists of
2008 MY vehicle designs, these analyses were focused on these vehicles,
their technology and their CO2 emission levels.
Comparing CO2 emissions across manufacturers is not a
simple task. In addition to widely varying vehicle styles, designs, and
sizes, manufacturers have implemented fuel efficient technologies to
varying degrees, as indicated in Table III.D.6-1. The projected levels
of requisite technology to enable compliance with the final 2016
standards shown in Table III.D.6-3 account for two of the major factors
which can affect CO2 emissions (1) Level of technology
already being utilized and (2) vehicle size, as represented by
footprint.
For example, the fuel economy of a manufacturer's 2008 vehicles may
be relatively high because of the use of advanced technologies. This is
the case with Toyota's high sales of their Prius hybrid. However, the
presence of this technology in a 2008 vehicle eliminates the ability to
significantly reduce CO2 further through the use of this
technology. In the extreme, if a manufacturer were to hybridize a high
level of its sales in 2016, it does not matter whether this technology
was present in 2008 or whether it would be added in order to comply
with the standards. The final level of hybrid technology would be the
same. Thus, the level at which technology is present in 2008 vehicles
does not explain the difference in requisite technology levels shown in
Table III.D.6-3.
Similarly, the final CO2 emission standards adjust the
required CO2 level according to a vehicle's footprint,
requiring lower absolute emission levels from smaller vehicles. Thus,
just because a manufacturer produces larger vehicles than another
manufacturer does not explain the differences seen in Table III.D.6-3.
In order to remove these two factors from our comparison, the EPA
lumped parameter model described above was used to estimate the degree
to which technology present on each 2008 MY vehicle in our reference
fleet was improving fuel efficiency. The effect of this technology was
removed and each vehicle's CO2 emissions were estimated as
if it utilized no additional fuel efficiency technology beyond the
baseline. The differences in vehicle size were accounted for by
determining the difference between the sales-weighted average of each
manufacturer's ``no technology'' CO2 levels to their
required CO2 emission level under the final 2016 standards.
The industry-wide difference was subtracted from each manufacturer's
value to highlight which manufacturers had lower and higher than
average ``no technology'' emissions. The results are shown in Figure
III.D.6-1.
BILLING CODE 6560-50-P

[[Page 25458]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.019

BILLING CODE 6560-50-C

[[Page 25459]]

As can be seen in Table III.D.6-3 the manufacturers projected to
require the greatest levels of technology also show the highest offsets
relative to the industry. The greatest offset shown in Figure III.D.6-1
is for Tata's trucks (Land Rover). These vehicles are estimated to have
100 g/mi greater CO2 emissions than the average 2008 MY
truck after accounting for differences in the use of fuel saving
technology and footprint. The lowest adjustment is for Subaru's trucks,
which have 50 g/mi CO2 lower emissions than the average
truck.
While this comparison confirms the differences in the technology
penetrations shown in Table III.D.6-3, it does not yet explain why
these differences exist. Two well-known factors affecting vehicle fuel
efficiency are vehicle weight and acceleration performance (henceforth
referred to as ``performance''). The footprint-based form of the final
CO2 standard accounts for most of the difference in vehicle
weight seen in the 2008 MY fleet. However, even at the same footprint,
vehicles can have varying weights. Higher performing vehicles also tend
to have higher CO2 emissions over the two-cycle fuel economy
test procedure. So manufacturers with higher average performance levels
will tend to have higher average CO2 emissions for any given
footprint. This variability at any given footprint contributes to much
of the scatter in the data (shown for example on plots like Figures
II.C.1-3 through II.C.1-6).
We developed a methodology to assess the impact of these two
factors on each manufacturer's projected compliance with the 2016
standards. First, we had to remove (or isolate) the effect of
CO2 control technology already being employed on 2008
vehicles. As described above, 2008 vehicles exhibit a wide range of
control technology and leaving these impacts in place would confound
the assessment of performance and weight on CO2 emissions.
Thus, the first step was to estimate each vehicle's ``no technology''
CO2 emissions. To do this, we used the EPA lumped parameter
model (described in the TSD) to estimate the overall percentage
reduction in CO2 emissions associated with technology
already on the vehicle and then backed out this effect mathematically.
Second, we performed a least-square linear regression of these no
technology CO2 levels against curb weight and the ratio of
rated engine horsepower to curb weight simultaneously. The ratio of
rated engine horsepower to curb weight is a good surrogate for
acceleration performance and the data is available for all vehicles,
whereas the zero to sixty time is not. Both factors were found to be
statistically significant at the 95% confidence level. Together, they
explained over 80% of the variability in vehicles' CO2
emissions for cars and over 70% for trucks. Third, we determined the
sales-weighted average curb weight per footprint for cars and trucks,
respectively, for the fleet as a whole. We also determined the sales-
weighted average of the ratio of rated engine horsepower to curb weight
for cars and trucks, respectively, for the fleet as a whole. Fourth, we
adjusted each vehicle's ``no technology'' CO2 emissions to
eliminate the degree to which the vehicle had higher or lower
acceleration performance or curb weight per footprint relative to the
car or truck fleet as a whole. For example, if a car's ratio of
horsepower to weight was 0.007 and the average ratio for all cars was
0.006, then the vehicle's ``no technology'' CO2 emission
level was reduced by the difference between these two values (0.001)
times the impact of the ratio of horsepower to weight on car
CO2 emissions from the above linear regression. Finally, we
substituted these performance and weight adjusted CO2
emission levels for the original, ``no technology'' CO2
emission levels shown in Figure III.D.6-1. The results are shown in
Figure III.D.6-2.
BILLING CODE 6560-50-P

[[Page 25460]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.020

BILLING CODE 6560-50-C

[[Page 25461]]

First, note that the scale in Figure III.D.6-2 is much smaller by a
factor of 3 than that in Figure III.D.6-1. In other words, accounting
for differences in vehicle weight (at constant footprint) and
performance dramatically reduces the variability among the
manufacturers' CO2 emissions. Most of the manufacturers with
high positive offsets in Figure III.D.6-1 now show low or negative
offsets. For example, BMW's and VW's trucks show very low
CO2 emissions. Tata's emissions are very close to the
industry average. Daimler's vehicles are no more than 10 g/mi above the
average for the industry. This analysis indicates that the primary
reasons for the differences in technology penetrations shown for the
various manufacturers in Table III.D.6-3 are weight and acceleration
performance. EPA has not determined why some manufacturers' vehicle
weight is relatively high for its footprint value, or whether this
weight provides additional utility for the consumer. Performance is
more straightforward. Some consumers desire high-acceleration
performance and some manufacturers orient their sales towards these
consumers. However, the cost in terms of CO2 emissions is
clear. Manufacturers producing relatively heavy or high performance
vehicles presently (with concomitant increased CO2
emissions) will require greater levels of technology in order to meet
the final CO2 standards in 2016.
As can be seen from Table III.D.6-3 above, widespread use of
several technologies is projected due to the final standards. The vast
majority of engines are projected to be converted to direct injection,
with some of these engines including cylinder deactivation or
turbocharging and downsizing. More than 60 percent of all transmissions
are projected to be either 6+ speed automatic transmissions or dual-
clutch automated manual transmissions. More than one-third of the fleet
is projected to be equipped with 42 volt start-stop capability. This
technology was not utilized in 2008 vehicles, but as discussed above,
promises significant fuel efficiency improvement at a moderate cost.
In their comments, Porsche stated that their vehicles have twice
the power-to-weight ratio as the fleet average and that their vehicles
presently have a high degree of technology penetration, which allows
them to meet the 2009 CAFE standards. Porsche also commented that the
2016 standards are not feasible for their firm, in part due to the high
level of technologies already present in their vehicles and due to
their ``very long production life cycles''. BMW in their comments
stated that their vehicles are ``feature-dense'' thus ``requiring
additional efforts to comply'' with future standards.\258\ Ferrari, in
their comments, states that the standards are not feasible for high-
performance sports cars without compromising on their
``distinctiveness''. They also state that because they already have
many technologies on the vehicles, ``there are limited possibilities
for further improvements.'' Finally Ferrari states that smaller volume
manufacturers have higher costs ``because they can be distributed over
very limited production volumes'', and they have longer product
lifecycles. The latter view was also shared by Lotus. These comments
will be addressed below, but are cited here as supporting the
conclusions from the above analysis that high-performance and feature-
dense vehicles have a greater challenge meeting the 2016 standards. In
general, other manufacturers covering the rest of the fleet and other
commenters agreed with EPA's analysis in the proposal of projected
technology usage, and supported the view that the 2016 model year
standards were feasible in the lead-time provided.
---------------------------------------------------------------------------

\258\ As a side note, one of the benefits for the off-cycle
technology credits allowed in this final rule is the opportunity
this flexibility provides for some of these `feature-dense' vehicles
to generate such credits to assist, to some extent, in the
companies' ability to comply.
---------------------------------------------------------------------------

In response to the comments above, EPA foresees no significant
technical or engineering issues with the projected deployment of these
technologies across the fleet by MY 2016, with their incorporation
being folded into the vehicle redesign process (with the exception of
some of the small volume manufacturers). All of these technologies are
commercially available now. The automotive industry has already begun
to convert its port fuel-injected gasoline engines to direct injection.
Cylinder deactivation and turbocharging technologies are already
commercially available. As indicated in Table III.D.6-1, high-speed
transmissions are already widely used. However, while more common in
Europe, automated manual transmissions are not currently used
extensively in the U.S. Widespread use of this technology would require
significant capital investment but does not present any significant
technical or engineering issues. Start-stop systems based on a 42-volt
architecture also represent a challenge because of the complications
involved in a changeover to a higher voltage electrical architecture.
However, with appropriate capital investments (which are captured in
the EPA estimated costs), these technology penetration rates are
achievable within the timeframe of this rule. While most manufacturers
have some plans for these systems, our projections indicate that their
use may exceed 35% of sales, with some manufacturers projected to use
higher levels.
Most manufacturers are not projected to hybridize any vehicles to
comply with the final standards. The hybrids shown for Toyota are
projected to be sold even in the absence of the final standards.
However the relatively high hybrid penetrations (14-15%) projected for
BMW, Daimler, Porsche, Tata and Volkswagen deserve further discussion.
These manufacturers are all projected by the OMEGA model to utilize the
maximum application of full hybrids allowed by our model in this
timeframe, which is 15 percent.
As discussed in the EPA RIA, a maximum 2016 technology penetration
rate of 85% is projected for the vast majority of available
technologies, however, for full hybrid systems the projection shows
that given the available lead-time full hybrids can only be applied to
approximately 15% of a manufacturer's fleet. This number of course can
vary by manufacturer. Hybrids are a relatively costly technology option
which requires significant changes to a vehicle's powertrain design,
and EPA estimates that manufacturers will require a significant amount
of lead time and capital investment to introduce this technology into
the market in very large numbers. Thus the EPA captures this
significant change in production facilities with a lower penetration
cap. A more thorough discussion of lead time limitations can be found
below and in Section III.B.5.
While the hybridization levels of BMW, Daimler, Porsche, Tata and
Volkswagen are relatively high, the sales levels of these five
manufacturers are relatively low. Thus, industry-wide, hybridization
reaches only 4 percent, compared with 3 percent in the reference case.
This 4 percent level is believed to be well within the capability of
the hybrid component industry by 2016. Thus, the primary challenge for
these five companies would be at the manufacturer level, redesigning a
relatively large percentage of sales to include hybrid technology. The
final TLAAS provisions will provide significant needed lead time to
these manufacturers for pre-2016 compliance, since all qualified
companies are able to take advantage of these provisions.
By 2016, it is likely that these manufacturers would also be able
to

[[Page 25462]]

change vehicle characteristics which currently cause their vehicles to
emit much more CO2 than similar sized vehicles produced by
other manufacturers. These factors may include changes in model mix,
further mass reduction, electric and/or plug-in hybrid vehicles as well
as technologies that may not be included in our packages. Also,
companies may have technology penetration rates of less costly
technologies (listed in the above tables) greater than 85%, and they
may also be able to apply hybrid technology to more than 15 percent of
their fleet (while the 15% cap on the application of hybrid technology
is reasonable for the industry as a whole, higher percentages are
certainly possible for individual manufacturers, particularly those
with small volumes). For example, a switch to a low GWP alternative
refrigerant in a large fraction of a fleet can replace many other much
more costly technologies, but this option is not captured in the
modeling. In addition, these manufacturers can also take advantage of
flexibilities, such as early credits for air conditioning and trading
with other manufacturers.
EPA believes it is likely that there will be certain high volume
manufacturers that will earn a significant amount of early GHG credits
starting in 2010 that would expire 5 years later, by 2015, unused. It
is possible that these manufacturers may be willing to sell these
credits to manufacturers with whom there is little or no direct
competition.\259\ Furthermore, a large number of manufacturers have
also stated publicly that they support the 2016 standards. The
following companies have all submitted letters in support of the
national program, including the 2016 MY levels discussed above: BMW,
Chrysler, Daimler, Ford, GM, Nissan, Honda, Mazda, Toyota, and
Volkswagen. This supports the view that the emissions reductions needed
to achieve the standards are technically and economically feasible for
all these companies, and that EPA's projection of model year 2016 non-
compliance for BMW, Daimler, and Volkswagen is based on an inability of
our model at this time to fully account for the full flexibilities of
the EPA program as well as the potentially unique technology approaches
or new product offerings which these manufacturers are likely to
employ.
---------------------------------------------------------------------------

\259\ For example, a manufacturer that only sells electric
vehicles may very well sell the credits they earn to another
manufacturer that does not sell any electric vehicles.
---------------------------------------------------------------------------

In addition, manufacturers do not need to apply technology exactly
according to our projections. Our projections simply indicate one path
which would achieve compliance. Those manufacturers whose vehicles are
heavier (feature dense) and higher performing than average in
particular have additional options to facilitate compliance and reduce
their technological burden closer to the industry average. These
options include decreasing the mass of the vehicles and/or decreasing
the power output of the engines. Finally, EPA allows compliance to be
shown through the use of emission credits obtained from other
manufacturers. Especially for the lower volume sales of some
manufacturers that could be one component of an effective compliance
strategy, reducing the technology that needs to be employed on their
vehicles.
For light-duty cars and trucks, manufacturers have available to
them a range of technologies that are currently commercially available
and can feasibly be employed in their vehicles by MY 2016. Our modeling
projects widespread use of these technologies as a technologically
feasible approach to complying with the final standards. Comments from
the manufacturers provided broad support for this conclusion. A limited
number of commenters presented specific concerns about their technology
opportunities, and EPA has described above (and elsewhere in the rule)
the paths available for them to comply.
In sum, EPA believes that the emissions reductions called for by
the final standards are technologically feasible, based on projections
of widespread use of commercially available technology, as well as use
by some manufacturers of other technology approaches and compliance
flexibilities not fully reflected in our modeling.
EPA also projected the cost associated with these projections of
technology penetration. Table III.D.6-4 shows the cost of technology in
order for manufacturers to comply with the 2011 MY CAFE standards, as
well as those associated with the final 2016 CO2 emission
standards. The latter costs are incremental to those associated with
the 2011 MY standards and also include $60 per vehicle, on average, for
the cost of projected use of improved air-conditioning systems.\260\
---------------------------------------------------------------------------

\260\ Note that the actual cost of the A/C technology is
estimated at $71 per vehicle as shown in Table III.D.2-3. However,
we expect only 85 percent of the fleet to add that technology.
Therefore, the cost of the technology when spread across the entire
fleet is $60 per vehicle ($71 x 85% = $60).

Table III.D.6-4--Cost of Technology per Vehicle in 2016 ($2007)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2011 MY CAFE standards, relative to 2008 MY Final 2016 CO2 standards, relative to 2011 MY
------------------------------------------------ CAFE standards
-----------------------------------------------
Cars Trucks All Cars Trucks All
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW..................................................... $346 $423 $368 $1,558 $1,195 $1,453
Chrysler................................................ 33 116 77 1,129 1,501 1,329
Daimler................................................. 468 683 536 1,536 931 1,343
Ford.................................................... 73 161 106 1,108 1,442 1,231
General Motors.......................................... 31 181 102 899 1,581 1,219
Honda................................................... 0 0 0 635 473 575
Hyundai................................................. 0 69 10 802 425 745
Kia..................................................... 0 42 7 667 247 594
Mazda................................................... 0 0 0 855 537 808
Mitsubishi.............................................. 328 246 295 817 1,218 978
Nissan.................................................. 0 61 18 686 1,119 810
Porsche................................................. 473 706 550 1,506 759 1,257
Subaru.................................................. 68 62 66 962 790 899
Suzuki.................................................. 49 232 79 1,015 537 937
Tata.................................................... 611 1,205 845 1,181 680 984
Toyota.................................................. 0 0 0 381 609 455
Volkswagen.............................................. 228 482 272 1,848 972 1,694

[[Page 25463]]

Overall................................................. 63 138 89 870 1,099 948
--------------------------------------------------------------------------------------------------------------------------------------------------------

As can be seen, the industry average cost of complying with the
2011 MY CAFE standards is quite low, $89 per vehicle. This cost is $11
per vehicle higher than that projected in the NPRM. This change is very
small and is due to several factors, mainly changes in the projected
sales of each manufacturer's specific vehicles, and changes in
estimated technology costs. Similar to the costs projected in the NPRM,
the range of costs across manufacturers is quite large. Honda, Mazda
and Toyota are projected to face no cost. In contrast, Mitsubishi,
Porsche, Tata and Volkswagen face costs of at least $272 per vehicle.
As described above, three of these last four manufacturers (all but
Mitsubishi) face high costs to meet even the 2011 MY CAFE standards due
to either their vehicles' weight per unit footprint or performance.
Porsche would have been projected to face lower costs in 2016 if they
were not expected to pay CAFE fines in 2011.
As shown in the last row of Table III.D.6-4, the average cost of
technology to meet the final 2016 standards for cars and trucks
combined relative to the 2011 MY CAFE standards is $948 per vehicle.
This is $103 lower than that projected in the NPRM, due primarily to
lower technology cost projections for the final rule compared to the
NPRM for certain technologies. (See Chapter 1 of the Joint TSD for a
detailed description of how our technology costs for the final rule
differ from those used in the NPRM). As was the case in the NPRM, Table
III.D.6-4 shows that the average cost for cars would be slightly lower
than that for trucks. Toyota and Honda show projected costs
significantly below the average, while BMW, Porsche, Tata and
Volkswagen show significantly higher costs. On average, the $948 per
vehicle cost is significant, representing 3.4 percent of the total cost
of a new vehicle. However, as discussed below, the fuel savings
associated with the final standards exceed this cost significantly. In
general, commenters supported EPA's cost projections, as discussed in
Section II.
While the CO2 emission compliance modeling using the
OMEGA model focused on the final 2016 MY standards, the final standards
for 2012-2015 are also feasible. As discussed above, manufacturers
develop their future vehicle designs with several model years in view.
Generally, the technology estimated above for 2016 MY vehicles
represents the technology which would be added to those vehicles which
are being redesigned in 2012-2015. The final CO2 standards
for 2012-2016 reduce CO2 emissions at a fairly steady rate.
Thus, manufacturers which redesign their vehicles at a fairly steady
rate will automatically comply with the interim standard as they plan
for compliance in 2016.
Manufacturers which redesign much fewer than 20% of their sales in
the early years of the final program would face a more difficult
challenge, as simply implementing the ``2016 MY'' technology as
vehicles are redesigned may not enable compliance in the early years.
However, even in this case, manufacturers would have several options to
enable compliance. One, they could utilize the debit carry-forward
provisions described above. This may be sufficient alone to enable
compliance through the 2012-2016 MY time period, if their redesign
schedule exceeds 20% per year prior to 2016. If not, at some point, the
manufacturer might need to increase their use of technology beyond that
projected above in order to generate the credits necessary to balance
the accrued debits. For most manufacturers representing the vast
majority of U.S. sales, this would simply mean extending the same
technology to a greater percentage of sales. The added cost of this in
the later years of the program would be balanced by lower costs in the
earlier years. Two, the manufacture could take advantage of the many
optional credit generation provisions contained in this final rule,
including early-credit generation for model years 2009-2011, credits
for advanced technology vehicles, and credits for the application of
technology which result in off-cycle GHG reductions. Finally, the
manufacturer could buy credits from another manufacturer. As indicated
above, several manufacturers are projected to require less stringent
technology than the average. These manufacturers would be in a position
to provide credits at a reasonable technology cost. Thus, EPA believes
the final standards for 2012-2016 would be feasible. Further discussion
of the technical feasibility of the interim year standards, including
for smaller volume manufacturers can be found in Section III.B, in the
discussion on the Temporary Leadtime Allowance Alternative Standards.
7. What other fleet-wide CO2 levels were considered?
Two alternative sets of CO2 standards were considered.
One set would reduce CO2 emissions at a rate of 4 percent
per year. The second set would reduce CO2 emissions at a
rate of 6 percent per year. The analysis of these standards followed
the exact same process as described above for the final standards. The
only difference was the level of CO2 emission standards. The
footprint-based standard coefficients of the car and truck curves for
these two alternative control scenarios were discussed above. The
resultant projected CO2 standards in 2016 for each
manufacturer under these two alternative scenarios and under the final
rule are shown in Table III.D.7-1.

Table III.D.7-1--Overall Average CO2 Emission Standards by Manufacturer in 2016
----------------------------------------------------------------------------------------------------------------
4% per year Final Rule 6% per year
----------------------------------------------------------------------------------------------------------------
BMW....................................................... 248 244 224
Chrysler.................................................. 270 266 245
Daimler................................................... 260 256 236
Ford...................................................... 261 257 237
General Motors............................................ 275 271 250
Honda..................................................... 248 244 224

[[Page 25464]]

Hyundai................................................... 234 231 212
Kia....................................................... 239 236 217
Mazda..................................................... 232 228 210
Mitsubishi................................................ 244 239 219
Nissan.................................................... 250 245 226
Porsche................................................... 237 233 213
Subaru.................................................... 238 234 214
Suzuki.................................................... 222 218 199
Tata...................................................... 263 259 239
Toyota.................................................... 249 245 225
Volkswagen................................................ 236 232 213
Overall................................................... 254 250 230
----------------------------------------------------------------------------------------------------------------

Tables III.D.7-2 and III.D.7-3 show the technology penetration
levels for the 4 percent per year and 6 percent per year standards in
2016.

Table III.D.7-2--Technology Penetration--4% per Year CO2 Standards in 2016: Cars and Trucks Combined
[In percent]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dual Mass
GDI OHC-DEAC Turbo Diesel 6 Speed clutch Start-stop Hybrid reduction
auto trans trans (%)
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW......................................... 80 21 61 6 13 63 65 14 5
Chrysler.................................... 67 13 17 0 26 52 54 0 6
Daimler *................................... 76 30 53 5 12 72 67 14 5
Ford........................................ 77 18 16 0 25 58 59 0 5
General Motors.............................. 62 24 11 0 7 57 57 0 5
Honda....................................... 44 6 2 0 0 49 15 2 2
Hyundai..................................... 52 0 1 0 3 52 28 0 3
Kia......................................... 37 0 1 0 0 57 0 0 2
Mazda....................................... 79 0 14 1 17 66 60 0 5
Mitsubishi.................................. 85 0 31 0 16 72 72 0 6
Nissan...................................... 69 7 11 0 2 64 61 1 6
Porsche *................................... 83 15 62 8 5 45 62 15 4
Subaru...................................... 72 0 9 0 0 70 37 0 3
Suzuki...................................... 70 0 0 0 3 67 67 0 3
Tata *...................................... 85 55 27 0 14 70 70 15 5
Toyota...................................... 15 7 0 0 13 30 7 12 1
Volkswagen *................................ 82 18 71 11 10 68 60 15 4
Overall..................................... 56 13 14 1 11 53 41 4 4
Increase over 2011 CAFE..................... 46 11 7 1 -5 46 38 2 4
--------------------------------------------------------------------------------------------------------------------------------------------------------
\*\ These manufacturers were unable to meet the final 2016 standards with the imposed caps on technology.

Table III.D.7-3--Technology Penetration--6% per Year Alternative Standards in 2016: Cars and Trucks Combined
[In percent]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dual Mass
GDI OHC-DEAC Turbo Diesel 6 Speed clutch Start-stop Hybrid reduction
auto trans trans (%)
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW *....................................... 80 21 61 6 13 63 65 14 5
Chrysler.................................... 85 13 50 0 3 82 83 2 8
Daimler *................................... 76 30 53 5 12 72 67 14 5
Ford*....................................... 85 13 57 0 4 74 75 10 7
General Motors.............................. 85 25 43 0 2 83 83 2 8
Honda....................................... 68 6 10 0 1 65 65 2 6
Hyundai..................................... 73 1 12 0 9 64 64 0 5
Kia......................................... 62 0 1 0 0 62 61 0 5
Mazda....................................... 85 0 19 1 4 80 82 0 7
Mitsubishi *................................ 85 4 42 0 4 75 75 10 7
Nissan...................................... 85 8 38 0 0 78 81 4 8
Porsche *................................... 83 15 62 8 5 45 62 15 4
Subaru...................................... 84 0 18 1 3 79 80 0 6
Suzuki...................................... 85 0 85 0 0 85 85 0 8
Tata *...................................... 85 55 27 0 14 70 70 15 5
Toyota...................................... 71 7 5 0 20 49 47 12 4

[[Page 25465]]

Volkswagen *................................ 82 18 71 11 10 68 60 15 4
Overall..................................... 79 12 33 1 7 69 69 6 6
Increase over 2011 CAFE..................... 69 10 26 1 -9 62 66 4 6
--------------------------------------------------------------------------------------------------------------------------------------------------------
* These manufacturers were unable to meet the final 2016 standards with the imposed caps on technology.

With respect to the 4 percent per year standards, the levels of
requisite control technology are lower than those under the final
standards, as would be expected. Industry-wide, the largest decreases
were a 7 percent decrease in use of gasoline direct injection engines,
a 4 percent decrease in the use of dual clutch transmissions, and a 2
percent decrease in the application of start-stop technology. On a
manufacturer specific basis, the most significant decreases were a 10
percent or larger decrease in the use of stop-start technology for
Honda, Kia, Mitsubishi and Suzuki and a 12 percent drop in turbocharger
use for Mitsubishi. These are relatively small changes and are due to
the fact that the 4 percent per year standards only require 4 g/mi
CO2 less control than the final standards in 2016. Porsche,
Tata and Volkswagen continue to be unable to comply with the
CO2 standards in 2016, even under the 4 percent per year
standard scenario. BMW just complied under this scenario, so its costs
and technology penetrations are the same as under the final standards.
With respect to the 6 percent per year standards, the levels of
requisite control technology increased substantially relative to those
under the final standards, as again would be expected. Industry-wide,
the largest increase was a 25 percent increase in the application of
start-stop technology and 13-17 percent increases in the use of
gasoline direct injection engines, turbocharging and dual clutch
transmissions. On a manufacturer specific basis, the most significant
increases were a 10 percent increase in hybrid penetration for Ford and
Mitsubishi. These are more significant changes and are due to the fact
that the 6 percent per year standards require 20 g/mi CO2
more control than the final standards in 2016. Our projections for BMW,
Porsche, Tata and Volkswagen continue to show they are unable to comply
with the CO2 standards in 2016, so our projections for these
manufacturers do not differ relative to the final standards, though the
amount of short-fall for each firm increases significantly, by an
additional 20 g/mi CO2 per firm. However, Ford and
Mitsubishi join this list as can be seen from Figure III.D.6-2. The
CO2 emissions from Ford's cars are very similar to those of
the industry when adjusted for technology, weight and performance.
However, their trucks emit more than 25% more CO2 per mile
than the industry average. It is possible that addressing this issue
would resolve their difficulty in complying with the 6 percent per year
scenario. Both Mitsubishi's cars and truck emit roughly 10% more than
the industry average vehicles after adjusting for technology, weight
and performance. Again, addressing this issue could resolve their
difficulty in complying with the 6 percent per year scenario. Five
manufacturers are projected to need to increase their use of start-stop
technology by at least 30 percent.
Table III.D.7-4 shows the projected cost of the two alternative
sets of standards.

Table III.D.7-4--Technology Cost per Vehicle in 2016--Alternative Standards ($2007)
--------------------------------------------------------------------------------------------------------------------------------------------------------
4 Percent per year standards, relative to 2011 6 Percent per year standards, relative to 2011
MY CAFE standards MY CAFE standards
-----------------------------------------------------------------------------------------------
Cars Trucks All Cars Trucks All
--------------------------------------------------------------------------------------------------------------------------------------------------------
BMW..................................................... $1,558 $1,195 $1,453 $1,558 $1,195 $1,453
Chrysler................................................ 1,111 1,236 1,178 1,447 2,156 1,827
Daimler................................................. 1,536 931 1,343 1,536 931 1,343
Ford.................................................... 1,013 1,358 1,140 1,839 2,090 1,932
General Motors.......................................... 834 1,501 1,148 1,728 2,030 1,870
Honda................................................... 598 411 529 894 891 893
Hyundai................................................. 769 202 684 1,052 1,251 1,082
Kia..................................................... 588 238 527 1,132 247 979
Mazda................................................... 766 537 733 1,093 1,083 1,092
Mitsubishi.............................................. 733 1,164 906 1,224 1,840 1,471
Nissan.................................................. 572 1,119 729 1,151 1,693 1,306
Porsche................................................. 1,506 759 1,257 1,506 759 1,257
Subaru.................................................. 962 616 836 1,173 1,316 1,225
Suzuki.................................................. 1,015 179 879 1,426 1,352 1,414
Tata.................................................... 1,181 680 984 1,181 680 984
Toyota.................................................. 323 560 400 747 906 799
Volkswagen.............................................. 1,848 972 1,694 1,848 972 1,694
Overall................................................. 811 1,020 883 1,296 1,538 1,379
--------------------------------------------------------------------------------------------------------------------------------------------------------

[[Page 25466]]

As can be seen, the average cost of the 4 percent per year
standards is only $65 per vehicle less than that for the final
standards. This incremental cost is very similar to that projected in
the NPRM. In contrast, the average cost of the 6 percent per year
standards is over $430 per vehicle more than that for the final
standards, which is $80 less than that projected in the NPRM (again due
to lower technology costs). Compliance costs are entering the region of
non-linearity. The $65 cost savings of the 4 percent per year standards
relative to the final rule represents $19 per g/mi CO2
increase. The $430 cost increase of the 6 percent per year standards
relative to the final rule represents a 25 per g/mi CO2
increase. More importantly, two additional manufacturers, Ford and
Mitsubishi, are projected to be unable to comply with the 6% per year
standards. In addition, under the 6% per year standards, four
manufacturers (Chrysler, General Motors, Suzuki and Nissan) are within
2 g/mi CO2 of the minimum achievable levels projected by
EPA's OMEGA model analysis for 2016.
EPA does not believe the 4% per year alternative is an appropriate
standard for the MY 2012-2016 time frame. As discussed above, the 250
g/mi final rule is technologically feasible in this time frame at
reasonable costs, and provides higher GHG emission reductions at a
modest cost increase over the 4% per year alternative (less than $100
per vehicle). In addition, the 4% per year alternative does not result
in a harmonized National Program for the country. Based on California's
letter of May 18, 2009, the emission standards under this alternative
would not result in the State of California revising its regulations
such that compliance with EPA's GHG standards would be deemed to be in
compliance with California's GHG standards for these model years. Thus,
the consequence of promulgating a 4% per year standard would be to
require manufacturers to produce two vehicle fleets: A fleet meeting
the 4% per year Federal standard, and a separate fleet meeting the more
stringent California standard for sale in California and the section
177 states. This further increases the costs of the 4% per year
standard and could lead to additional difficulties for the already
stressed automotive industry.
EPA also does not believe the 6% per year alternative is an
appropriate standard for the MY 2012-2016 time frame. As shown in
Tables III.D.7-3 and III.D.7-4, the 6% per year alternative represents
a significant increase in both the technology required and the overall
costs compared to the final standards. In absolute percent increases in
the technology penetration, compared to the final standards the 6% per
year alternative requires for the industry as a whole: An 18% increase
in GDI fuel systems, an 11% increase in turbo-downsize systems, a 6%
increase in dual-clutch automated manual transmissions (DCT), and a 9%
increase in start-stop systems. For a number of manufacturers the
expected increase in technology is greater: For GM, a 15% increase in
both DCTs and start-stop systems, for Nissan a 9% increase in full
hybrid systems, for Ford an 11% increase in full hybrid systems, for
Chrysler a 34% increase in both DCT and start-stop systems and for
Hyundai a 23% increase in the overall penetration of DCT and start-stop
systems. For the industry as a whole, the per-vehicle cost increase for
the 6% per year alternative is nearly $500. On average this is a 50%
increase in costs compared to the final standards. At the same time,
CO2 emissions would be reduced by about 8%, compared to the
250 g/mi target level.
As noted above, EPA's OMEGA model predicts that for model year
2016, Ford, Mitsubishi, Mercedes, BMW, Volkswagen, Jaguar-Land Rover,
and Porsche do not meet their target under the 6 percent per year
scenario. In addition, Chrysler, General Motors, Suzuki and Nissan all
are within 2 grams/mi CO2 of maximizing the applicable
technology allowed under EPA's OMEGA model--that is, these companies
have almost no head-room for compliance. In total, these 11 companies
represent more than 58 percent of total 2016 projected U.S. light-duty
vehicle sales. This provides a strong indication that the 6 percent per
year standard is much more stringent than the final standards, and
presents a significant risk of non-compliance for many firms, including
four of the seven largest firms by U.S. sales.
These technology and cost increases are significant, given the
amount of lead-time between now and model years 2012-2016. In order to
achieve the levels of technology penetration for the final standards,
the industry needs to invest significant capital and product
development resources right away, in particular for the 2012 and 2013
model year, which is only 2-3 years from now. For the 2014-2016 time
frame, significant product development and capital investments will
need to occur over the next 2-3 years in order to be ready for
launching these new products for those model years. Thus a major part
of the required capital and resource investment will need to occur now
and over the next few years, under the final standards. EPA believes
that the final rule (a target of 250 gram/mile in 2016) already
requires significant investment and product development costs for the
industry, focused on the next few years.
It is important to note, and as discussed later in this preamble,
as well as in the Joint Technical Support Document and the EPA
Regulatory Impact Analysis document, the average model year 2016 per-
vehicle cost increase of nearly $500 includes an estimate of both the
increase in capital investments by the auto companies and the suppliers
as well as the increase in product development costs. These costs can
be significant, especially as they must occur over the next 2-3 years.
Both the domestic and transplant auto firms, as well as the domestic
and world-wide automotive supplier base, is experiencing one of the
most difficult markets in the U.S. and internationally that has been
seen in the past 30 years. One major impact of the global downturn in
the automotive industry and certainly in the U.S. is the significant
reduction in product development engineers and staffs, as well as a
tightening of the credit markets which allow auto firms and suppliers
to make the near-term capital investments necessary to bring new
technology into production. The 6% per year alternative standard would
impose significantly increased pressure on capital and other resources,
indicating it is too stringent for this time frame, given both the
relatively limited amount of lead-time between now and model years
2012-2016, the need for much of these resources over the next few
years, as well the current financial and related circumstances of the
automotive industry. EPA is not concluding that the 6% per year
alternative standards are technologically infeasible, but EPA believes
such standards for this time frame would be overly stringent given the
significant strain it would place on the resources of the industry
under current conditions. EPA believes this degree of stringency is not
warranted at this time. Therefore EPA does not believe the 6% per year
alternative would be an appropriate balance of various relevant factors
for model years 2012-1016.
Jaguar/Land Rover, in their comments, agreed that the more
stringent standards would not be economically practicable, and several
automotive firms indicated that the proposed standards, while feasible,
would be overly challenging.\261\ On the other hand, the Center for
Biological Diversity (henceforth referred to here as CBD), strongly
urged EPA to adopt more

[[Page 25467]]

stringent standards. CBD gives examples of higher standards in other
nations to support their contention that the standards should be more
stringent. CBD also claims that the agencies are ``setting standards
that deliberately delay implementation of technology that is available
now'' by setting lead time for the rule greater than 18 months. CBD
also accuses the agencies of arbitrarily ``adhering to strict five-year
manufacturer `redesign cycles.' '' CBD notes that the agencies have
stated that all of the ``technologies are already available today,''
and EPA and NHTSA's assessment is that manufacturers ``would be able to
meet the proposed standards through more widespread use of these
technologies across the fleet.'' Based on the agencies' previous
statements, CBD concludes that the fleet can meet the 250 g/mi target
in 2010. EPA believes that in all cases, CBD's analysis for feasibility
and necessary lead time is flawed.
---------------------------------------------------------------------------

\261\ See comments from Toyota, General Motors.
---------------------------------------------------------------------------

Other countries' absolute fleetwide standards are not a reliable or
directly relevant comparison. The fleet make-up in other nations is
quite different than that of the United States. CBD primarily cites the
European Union and Japan as examples. Both of these regions have a
large fraction of small vehicles (with lower average weight, and
footprint size) when compared to vehicles in the U.S. Also the U.S. has
a much greater fraction of light-duty trucks. In particular in Europe,
there is a much higher fraction of diesel vehicles in the existing
fleet, which leads to lower CO2 emissions in the baseline
fleet as compared to the U.S. This is in large part due to the
significantly different fuel prices seen in Europe as compared to the
U.S. The European fleet also has a much higher penetration of manual
transmission than the U.S., which also results in lower CO2
emissions. Moreover, these countries use different test cycles, which
bias CO2 emissions relative to the EPA 2 cycle test cycles.
When looked at from a technology-basis, with the exception of the
existing large penetration of diesels and manual transmissions in the
European fleet--there is no ``magic'' in the European and Japanese
markets which leads to lower fleet-wide CO2 emissions. In
fact, from a technology perspective, the standards contained in this
final rule are premised to a large degree on the same technologies
which the European and Japanese governments have relied upon to
establish their CO2 and fuel economy limits for this same
time frame and for the fleet mixes in their countries. That is for
example, large increases in the use of 6+ speed transmissions,
automated manual transmissions, gasoline direct injection, engine
downsizing and turbocharging, and start-stop systems. CBD has not
provided any detailed analysis of what technologies are available in
Europe which EPA is not considering--and there are no such ``magic''
technologies. The vast majority of the differences between the current
and future CO2 performance of the Japanese and European
light-duty vehicle fleets are due to differences in the size and
current composition of the vehicle fleets in those two regions--not
because EPA has ignored technologies which are available for
application to the U.S. market in the 2012-2016 time frame.
If CBD is advocating a radical reshifting of domestic fleet
composition, (such as requiring U.S. consumers to purchase much smaller
vehicles and requiring U.S. consumers to purchase vehicles with manual
transmissions), it is sufficient to say that standards forcing such a
result are not compelled under section 202(a), where reasonable
preservation of consumer choice remains a pertinent factor for EPA to
consider in balancing the relevant statutory factors. See also
International Harvester (478 F. 2d at 640 (Administrator required to
consider issues of basic demand for new passenger vehicles in making
technical feasibility and lead time determinations). Thus EPA believes
that the standard is at the proper level of stringency for the
projected domestic fleet in the 2012-2016 model years taking into
account the wide variety of consumer choice that is reflected in this
projection of the domestic fleet.
As mentioned earlier (in III.D.4), CBD's comments on available lead
time also are inaccurate. Under section 202(a), standards are to take
effect only ``after providing such period as the Administrator finds
necessary to permit the development and application of the requisite
technology, giving appropriate consideration to the cost of compliance
within such period.'' Having sufficient lead time includes among other
things, the time required to certify vehicles. For example, model year
2012 vehicles will be tested and certified for the EPA within a short
time after the rule is finalized, and this can start as early as
calendar year 2010, for MY 2012 vehicles that can be produced in
calendar year 2011. In addition, these 2012 MY vehicles have already
been fully designed, with prototypes built several years earlier. It
takes several years to redesign a vehicle, and several more to design
an entirely new vehicle not based on an existing platform. Thus,
redesign cycles are an inextricable component of adequate lead time
under the Act. A full line manufacturer only has limited staffing and
financial resources to redesign vehicles, therefore the redesigns are
staggered throughout a multi-year period to optimize human
capital.\262\ Furthermore, redesigns require a significant outlay of
capital from the manufacturer. This includes research and development,
material and equipment purchasing, overhead, benefits, etc. These costs
are significant and are included in the cost estimates for the
technologies in this rule. Because of the manpower and financial
capital constraints, it would only be possible to redesign all the
vehicles across a manufacturer's line simultaneously if the
manufacturer has access to tremendous amounts of ready capital and an
unrealistically large engineering staff. However no major automotive
firm in the world has the capability to undertake such an effort, and
it is unlikely that the supplier basis could support such an effort if
it was required by all major automotive firms. Even if this unlikely
condition were possible, the large engineering staff would then have to
be downsized or work on the next redesign of the entire line another
few years later. This would have the effect of increasing the cost of
the vehicles.
---------------------------------------------------------------------------

\262\ See for example ``How Automakers Plan Their Products'',
Center for Automotive Research, July 2007.
---------------------------------------------------------------------------

There is much evidence to indicate that the average redesign cycle
in the industry is about 5 years.\263\ There are some manufacturers who
have longer cycles (such as smaller manufacturers described above), and
there are others who have shorter cycles for some of their products.
EPA believes that there are no full line manufacturers who can maintain
significant redesigns of vehicles (with relative large sales) in 1 or 2
years, and CBD has provided no evidence indicating this is technically
feasible. A complete redesign of the entire U.S. light-duty fleet by
model year 2012 is clearly infeasible, and EPA believes that several
model years additional lead time is necessary in order for the
manufacturers to meet the standards. The graduated increase in the
stringency of the standards from MYs 2012 through 2016 accounts for
this needed lead time.
---------------------------------------------------------------------------

\263\ See for example ``Car Wars 2010-2013, The U.S. automotive
product pipeline'', John Murphy, Research Analyst, Bank of America/
Merrill Lynch research paper, July 15, 2009.
---------------------------------------------------------------------------

There are other reasons that the fleet cannot meet the 250g/mi
CO2 target in 2012 (much less in 2010). The commenter
reasons that if technology is in use now--even if limited use--it can

[[Page 25468]]

be utilized across the fleet nearly immediately. This is not the case.
An immediate demand from original equipment manufacturers (OEMs) to
supply 100% of the fleet with these technologies in 2012 would cause
their suppliers to encounter the same lead time issues discussed above.
Suppliers have limited capacity to change their current production over
to the newer technologies quickly. Part of this reason is due to
engineering, cost and manpower constraints as described above, but
additionally, the suppliers face an issue of ``stranded capital''. This
is when the basic tooling and machines that produce the technologies in
question need to be replaced. If these tools and machines are replaced
before they near the end of their useful life, the suppliers are left
with ``stranded capital'' i.e., a significant financial loss because
they are replacing perfectly good equipment with newer equipment. This
situation can also occur for the OEMs. In an extreme example, a plant
that switches over from building port fuel injected gasoline engines to
building batteries and motors, will require a nearly complete retooling
of the plant. In a less extreme example, a plant that builds that same
engine and switches over to suddenly building smaller turbocharged
direct injection engines with starter alternators might have
significant retooling costs as well as stranded capital. Finally, it
takes a significant amount of time to retool a factory and smoothly
validate the tooling and processes to mass produce a replacement
technology. This is why most manufacturers do this process over time,
replacing equipment as they wear out. CBD has not accounted for any of
these considerations. EPA believes that attempting to force the types
of massive technology penetration needed in the early model years of
the standard to achieve the 2016 standards would be physically and cost
prohibitive.
A number of automotive firms and associations (including the
Alliance of Automobile Manufacturers, Mercedes, and Toyota) commented
that the standards during the early model years, in particular MY 2012,
are too stringent, and that a more linear phase-in of the standards
beginning with the MY 2011 CAFE standards and ending with the 250 gram/
mi proposed EPA projected fleet-wide level in MY 2016 is more
appropriate. In the May 19, 2009 Joint Notice of Intent, EPA and NHTSA
stated that the standards would have ``* * * a generally linear phase-
in from MY 2012 through to model year 2016.'' (74 FR 24008). The
Alliance of Automobile Manufacturers stated that the phase-in of the
standards is not linear, and they proposed a methodology for the CAFE
standards to be a linear progression from MY 2011 to MY 2016. The
California Air Resources Board commented that the proposed level of
stringency, including the EPA proposed standards for MY 2012-2015, were
appropriate and urged EPA to finalize the standards as proposed and not
reduce the stringency in the early model years as this would result in
a large loss of the GHG reductions from the National Program. EPA
agrees with the comments from CARB, and we have not reduced the
stringency of the program for the early model years. While some
automotive firms indicated a desire to see a linear transition from the
Model Year 2011 CAFE standards, our technology and cost analysis
indicates that our standards are appropriate for these interim years.
As shown in Section III.H of this final rule, the final standards
result in significant GHG reductions, including the reductions from MY
2012-2015, and at reasonable costs, providing appropriate lead time.
The automotive industry commenters did not point to a specific
technical issue with the standards, but rather their desire for a
linear phase-in from the existing 2011 CAFE standards.
In summary, the EPA believes that the MY 2012-2016 standards
finalized are feasible and that there are compelling reasons not to
adopt more stringent standards, based on a reasonable weighing of the
statutory factors, including available technology, its cost, and the
lead time necessary to permit its development and application. For
further discussion of these issues, see Chapter 4 of the RIA as well as
the response to comments.

E. Certification, Compliance, and Enforcement

1. Compliance Program Overview
This section describes EPA's comprehensive program to ensure
compliance with emission standards for carbon dioxide (CO2),
nitrous oxide (N2O), and methane (CH4), as
described in Section III.B. An effective compliance program is
essential to achieving the environmental and public health benefits
promised by these mobile source GHG standards. EPA's GHG compliance
program is designed around two overarching priorities: (1) To address
Clean Air Act (CAA) requirements and policy objectives; and (2) to
streamline the compliance process for both manufacturers and EPA by
building on existing practice wherever possible, and by structuring the
program such that manufacturers can use a single data set to satisfy
both the new GHG and Corporate Average Fuel Economy (CAFE) testing and
reporting requirements. The EPA and NHTSA programs recognize, and
replicate as closely as possible, the compliance protocols associated
with the existing CAA Tier 2 vehicle emission standards, and with CAFE
standards. The certification, testing, reporting, and associated
compliance activities closely track current practices and are thus
familiar to manufacturers. EPA already oversees testing, collects and
processes test data, and performs calculations to determine compliance
with both CAFE and CAA standards. Under this coordinated approach, the
compliance mechanisms for both programs are consistent and non-
duplicative.
Vehicle emission standards established under the CAA apply
throughout a vehicle's full useful life. Today's rule establishes fleet
average greenhouse gas standards where compliance with the fleet
average is determined based on the testing performed at time of
production, as with the current CAFE fleet average. EPA is also
establishing in-use standards that apply throughout a vehicle's useful
life, with the in-use standard determined by adding an adjustment
factor to the emission results used to calculate the fleet average.
EPA's program will thus not only assess compliance with the fleet
average standards described in Section III.B, but will also assess
compliance with the in-use standards. As it does now, EPA will use a
variety of compliance mechanisms to conduct these assessments,
including pre-production certification and post-production, in-use
monitoring once vehicles enter customer service. Specifically, EPA is
establishing a compliance program for the fleet average that utilizes
CAFE program protocols with respect to testing, a certification
procedure that operates in conjunction with the existing CAA Tier 2
certification procedures, and an assessment of compliance with the in-
use standards concurrent with existing EPA and manufacturer Tier 2
emission compliance testing programs. Under this compliance program
manufacturers will also be afforded numerous flexibilities to help
achieve compliance, both stemming from the program design itself in the
form of a manufacturer-specific CO2 fleet average standard,
as well as in various credit banking and trading opportunities, as
described in Section III.C. EPA received broad comment from regulated
industry and from the public interest community supporting this overall
compliance program structure.

[[Page 25469]]

The compliance program is outlined in further detail below.
2. Compliance With Fleet-Average CO2 Standards
Fleet average emission levels can only be determined when a
complete fleet profile becomes available at the close of the model
year. Therefore, EPA will determine compliance with the fleet average
CO2 standards when the model year closes out, as is
currently the protocol under EPA's Tier 2 program as well as under the
current CAFE program. The compliance determination will be based on
actual production figures for each model and on model-level emissions
data collected through testing over the course of the model year.
Manufacturers will submit this information to EPA in an end-of-year
report which is discussed in detail in Section III.E.5.h below.
Manufacturers currently conduct their CAFE testing over an entire
model year to maximize efficient use of testing and engineering
resources. Manufacturers submit their CAFE test results to EPA and EPA
conducts confirmatory fuel economy testing at its laboratory on a
subset of these vehicles under EPA's Part 600 regulations. EPA's
proposal to extend this approach to the GHG program received
overwhelming support from vehicle manufacturers. EPA is finalizing GHG
requirements under which manufacturers will continue to perform the
model-level testing currently required for CAFE fuel economy
performance and measure and report the CO2 values for all
tests conducted.\264\ Manufacturers will submit one data set in
satisfaction of both CAFE and GHG requirements such that EPA's program
will not impose additional timing or testing requirements on
manufacturers beyond that required by the CAFE program. For example,
manufacturers currently submit fuel economy test results at the
subconfiguration and configuration levels to satisfy CAFE requirements.
Now manufacturers will also submit CO2 values for the same
vehicles. Section III.E.3 discusses how this will be implemented in the
certification process.
---------------------------------------------------------------------------

\264\ As discussed in Section III.B.1, vehicle and fleet average
compliance will be based on a combination of CO2, HC, and
CO emissions. This is consistent with the carbon balance methodology
used to determine fuel consumption for the labeling and CAFE
programs. The final regulations account for these total carbon
emissions appropriately and refer to the sum of these emissions as
the ``carbon-related exhaust emissions'' (CREE). Although regulatory
text uses the more accurate term ``CREE'' to represent the
CO2-equivalent sum of carbon emissions, the term
CO2 is used as shorthand throughout Section III.E as a
more familiar term for most readers.
---------------------------------------------------------------------------

a. Compliance Determinations
As described in Section III.B above, the fleet average standards
will be determined on a manufacturer by manufacturer basis, separately
for cars and trucks, using the footprint attribute curves. EPA will
calculate the fleet average emission level using actual production
figures and, for each model type, CO2 emission test values
generated at the time of a manufacturer's CAFE testing. EPA will then
compare the actual fleet average to the manufacturer's footprint
standard to determine compliance, taking into consideration use of
averaging and credits.
Final determination of compliance with fleet average CO2
standards may not occur until several years after the close of the
model year due to the flexibilities of carry-forward and carry-back
credits and the remediation of deficits (see Section III.C). A failure
to meet the fleet average standard after credit opportunities have been
exhausted could ultimately result in penalties and injunctive orders
under the CAA as described in Section III.E.6 below.
EPA received considerable comment about the need for transparency
in its implementation of the greenhouse gas program and specifically
about the need for public access to information about Agency compliance
determinations. Many comments emphasized the importance of making
greenhouse gas compliance information publicly available to ensure such
transparency. EPA also received comment from industry about the need to
protect confidential business information. Both transparency and
protection of confidential information are longstanding EPA practices,
and both will remain priorities in EPA's implementation of the
greenhouse gas program. EPA periodically provides mobile source
emissions and fuel economy information to the public, for example
through the annual Compliance Report \265\ and Fuel Economy Trends
Report.\266\ As proposed, EPA plans to expand these reports to include
GHG performance and compliance trends information, such as annual
status of credit balances or debits, use of various credit programs,
attained fleet average emission levels compared with standards, and
final compliance status for a model year after credit reconciliation
occurs. EPA intends to regularly disseminate non-confidential, model-
level and fleet information for each manufacturer after the close of
the model year. EPA will reassess data release needs and opportunities
once the program is underway.
---------------------------------------------------------------------------

\265\ 2007 Progress Report Vehicle and Engine Compliance
Activities; EPA-420-R-08-011; October 2008. This document is
available electronically at http://www.epa.gov/otaq/about/
420r08011.pdf.
\266\ Light-Duty Automotive Technology and Fuel-Economy Trends:
1975 Through 2008; EPA-420-S-08-003; September 2008. This document
is available electronically at http://www.epa.gov/otaq/fetrends.htm.
---------------------------------------------------------------------------

Beyond transparency in reporting emissions data and compliance
status, EPA is concerned, as a matter of principle moving into a new
era of greenhouse gas control, that greenhouse gas reductions reported
for purposes of compliance with the standards adopted in this rule will
be reflected in the real world and not just as calculated fleet average
emission levels or measured certification test results. Therefore EPA
will pay close attention to technical details behind manufacturer
reports. For example, EPA intends to look closely at each
manufacturer's certification testing procedures, GHG calculation
procedures, and laboratory correlation with EPA's laboratory, and to
carefully review manufacturer pre-production, production, and in-use
testing programs. In addition, EPA plans to monitor GHG performance
through its own in-use surveillance program in the coming years. This
will ensure that the environmental benefits of the rule are achieved as
well as ensure a level playing field for all.
b. Required Minimum Testing for Fleet Average CO2
EPA received no public comment on provisions that would extend
current CAFE testing requirements and flexibilities to the GHG program,
and is finalizing as proposed minimum testing requirements for fleet
average CO2 determination. EPA will require and use the same
test data to determine a manufacturer's compliance with both the CAFE
standard and the fleet average CO2 emissions standard. CAFE
requires manufacturers to submit test data representing at least 90% of
the manufacturer's model year production, by configuration.\267\ The
CAFE testing covers the vast majority of models in a manufacturer's
fleet. Manufacturers industry-wide currently test more than 1,000
vehicles each year to meet this requirement. EPA believes this minimum
testing requirement is necessary and applicable for calculating
accurate CO2 fleet average emissions. Manufacturers may test
additional

[[Page 25470]]

vehicles, at their option. As described above, EPA will use the
emissions results from the model-level testing to calculate a
manufacturer's fleet average CO2 emissions and to determine
compliance with the CO2 fleet average standard.
---------------------------------------------------------------------------

\267\ See 40 CFR 600.010-08(d).
---------------------------------------------------------------------------

EPA will continue to allow certain testing flexibilities that exist
under the CAFE program. EPA has always permitted manufacturers some
ability to reduce their test burden in tradeoff for lower fuel economy
numbers. Specifically the practice of ``data substitution'' enables
manufacturers to apply fuel economy test values from a ``worst case''
configuration to other configurations in lieu of testing them. The
substituted values may only be applied to configurations that would be
expected to have better fuel economy and for which no actual test data
exist. EPA will continue to accept use of substituted data in the GHG
program, but only when the substituted data are also used for CAFE
purposes.
EPA regulations for CAFE testing permit the use of analytically
derived fuel economy data in lieu of conducting actual fuel economy
tests in certain situations.\268\ Analytically derived data are
generated mathematically using expressions determined by EPA and are
allowed on a limited basis when a manufacturer has not tested a
specific vehicle configuration. This has been done as a way to reduce
some of the testing burden on manufacturers without sacrificing
accuracy in fuel economy measurement. EPA has issued guidance that
provides details on analytically derived data and that specifies the
conditions when analytically derived fuel economy data may be used. EPA
will apply the same guidance to the GHG program and will allow any
analytically derived data used for CAFE to also satisfy the GHG data
reporting requirements. EPA will revise the terms in the current
equations for analytically derived fuel economy to specify them in
terms of CO2. Analytically derived CO2 data will
not be permitted for the Emission Data Vehicle representing a test
group for pre-production certification, only for the determination of
the model level test results used to determine actual fleet-average
CO2 levels.
---------------------------------------------------------------------------

\268\ 40 CFR 600.006-08(e).
---------------------------------------------------------------------------

EPA is retaining the definitions needed to determine CO2
levels of each model type (such as ``subconfiguration,''
``configuration,'' ``base level,'' etc.) as they are currently defined
in EPA's fuel economy regulations.
3. Vehicle Certification
CAA section 203(a)(1) prohibits manufacturers from introducing a
new motor vehicle into commerce unless the vehicle is covered by an
EPA-issued certificate of conformity. Section 206(a)(1) of the CAA
describes the requirements for EPA issuance of a certificate of
conformity, based on a demonstration of compliance with the emission
standards established by EPA under section 202 of the Act. The
certification demonstration requires emission testing, and must be done
for each model year.\269\
---------------------------------------------------------------------------

\269\ CAA section 206(a)(1).
---------------------------------------------------------------------------

Under Tier 2 and other EPA emission standard programs, vehicle
manufacturers certify a group of vehicles called a test group. A test
group typically includes multiple vehicle car lines and model types
that share critical emissions-related features.\270\ The manufacturer
generally selects and tests one vehicle to represent the entire test
group for certification purposes. The test vehicle is the one expected
to be the worst case for the emission standard at issue. Emission
results from the test vehicle are used to assign the test group to one
of several specified bins of emissions levels, identified in the Tier 2
rule, and this bin level becomes the in-use emissions standard for that
test group.\271\
---------------------------------------------------------------------------

\270\ The specific test group criteria are described in 40 CFR
86.1827-01, car lines and model types have the meaning given in 40
CFR 86.1803-01.
\271\ Initially in-use standards were different from the bin
level determined at certification as the useful life level. The
current in-use standards, however, are the same as the bin levels.
In all cases, the bin level, reflecting useful life levels, has been
used for determining compliance with the fleet average.
---------------------------------------------------------------------------

Since compliance with the Tier 2 fleet average depends on actual
test group sales volumes and bin levels, it is not possible to
determine compliance with the fleet average at the time the
manufacturer applies for and receives a certificate of conformity for a
test group. Instead, EPA requires the manufacturer to make a good faith
demonstration in the certification application that vehicles in the
test group will both (1) comply throughout their useful life with the
emissions bin assigned, and (2) contribute to fleet-wide compliance
with the Tier 2 average when the year is over. EPA issues a certificate
for the vehicles included in the test group based on this
demonstration, and includes a condition in the certificate that if the
manufacturer does not comply with the fleet average, then production
vehicles from that test group will be treated as not covered by the
certificate to the extent needed to bring the manufacturer's fleet
average into compliance with Tier 2.
The certification process often occurs several months prior to
production and manufacturer testing may occur months before the
certificate is issued. The certification process for the Tier 2 program
is an efficient way for manufacturers to conduct the needed testing
well in advance of certification, and to receive the needed
certificates in a time frame which allows for the orderly production of
vehicles. The use of a condition on the certificate has been an
effective way to ensure compliance with the Tier 2 fleet average.
EPA will similarly condition each certificate of conformity for the
GHG program upon a manufacturer's demonstration of compliance with the
manufacturer's fleet-wide average CO2 standard. The
following discussion explains how EPA will integrate the new GHG
vehicle certification program into the existing certification program.
a. Compliance Plans
In an effort to expedite the Tier 2 program certification process
and facilitate early resolution of any compliance related concerns, EPA
conducts annual reviews of each manufacturer's certification, in-use
compliance and fuel economy plans for upcoming model year vehicles. EPA
meets with each manufacturer individually, typically before the
manufacturer begins to submit applications for certification for the
new model year. Discussion topics include compliance plans for the
upcoming model year, any new product offerings/new technologies,
certification and/or testing issues, phase-in and/or ABT plans, and a
projection of potential EPA confirmatory test vehicles. EPA has been
conducting these compliance preview meetings for more than 10 years and
has found them to be very useful for both EPA and manufacturers.
Besides helping to expedite the certification process, certification
preview meetings provide an opportunity to resolve potential issues
before the process begins. The meetings give EPA an early opportunity
to assess a manufacturer's compliance strategy, which in turn enables
EPA to address any potential concerns before plans are finalized. The
early interaction reduces the likelihood of unforeseen issues occurring
during the actual certification of a test group which can result in the
delay or even termination of the certification process.
For the reasons discussed above, along with additional factors, EPA
believes it is appropriate for manufacturers to include their GHG
compliance plan information as part of

[[Page 25471]]

the new model year compliance preview process. This requirement is both
consistent with existing practice under Tier 2 and very similar to the
pre-model year report required under existing and new CAFE regulation.
Furthermore, in light of the production weighted fleet average program
design in which the final compliance determination cannot be made until
after the end of the model year, EPA believes it is especially
important for manufacturers to demonstrate that they have a credible
compliance plan prior to the beginning of certification.
Several commenters raised concerns about EPA's proposal for
requiring manufacturers to submit GHG compliance plans. AIAM stated
that EPA did not identify a clear purpose for the review of the plans,
criteria for evaluating the plans, or consequences if EPA found the
plans to be unacceptable. AIAM also expressed concern over the
appropriateness of requiring manufacturers to prepare regulatory
compliance plans in advance, since vicissitudes of the market and other
factors beyond a manufacturer's direct control may change over the
course of the year and affect the model year outcome. Finally, AIAM
commented that EPA should not attempt to take any enforcement action
based on an asserted inadequacy of a plan. The comments stated that
compliance should be determined only after the end of a model year and
the subsequent credit earning period. The Alliance commented that there
was an inconsistency between the proposed preamble language and the
regulatory language in 600.514-12(a)(2)(i). The preamble language
indicated that the compliance report should be submitted prior to the
beginning of the model year and prior to the certification of any test
group, while the regulatory language stated that the pre-model year
report must be submitted during the month of December. The Alliance
pointed out that if EPA wanted GHG compliance plan information before
the certification of any test groups, the regulatory language would
need to be corrected.
EPA understands that a manufacturer's plan may change over the
course of a model year and that compliance information manufacturers
present prior to the beginning of a new model year may not represent
the final compliance outcome. Rather, EPA views the compliance plan as
a manufacturer's good-faith projection of strategy for achieving
compliance with the greenhouse gas standard. It is not EPA's intent to
base compliance action solely on differences between projections in the
compliance plan and end of year results. EPA understands that
compliance with the GHG program will be determined at the end of the
model year after all appropriate credits have been taken into
consideration.
As stated earlier, a requirement to include GHG compliance
information in the new model year compliance preview meetings is
consistent with long standing EPA policy. The information will provide
EPA with an early overview of the manufacturer's GHG compliance plan
and allow EPA to make an early assessment as to possible issues,
questions, or concerns with the program in order to expedite the
certification process and help manufacturers better understand overall
compliance provisions of the GHG program. Therefore, EPA is finalizing
revisions to 40 CFR 600.514-12 which will require manufacturers to
submit a compliance plan to EPA prior to the beginning of the model
year and prior to the certification of any test group. The compliance
plan must, at a minimum, include a manufacturer's projected footprint
profile, projected total and model-level production volumes, projected
fleet average and model-level CO2 emission values, projected
fleet average CO2 standards and projected fleet average
CO2 credit status. In addition, EPA will expect the
compliance plan to explain the various credit, transfer and trading
options that will be used to comply with the standard, including the
amount of credit the manufacturer intends to generate for air
conditioning leakage, air conditioning efficiency, off-cycle
technology, and various early credit programs. The compliance plan
should also indicate how and when any deficits will be paid off through
accrual of future credits.
EPA has corrected the inconsistency between the proposed preamble
and regulatory language with respect to when the compliance report must
be submitted and what level of information detail it must contain. EPA
is finalizing revisions to 40 CFR 600.514-12 which require the
compliance plan to be submitted to EPA prior to the beginning of the
model year and prior to the certification of any test group. Today's
action will also finalize simplified reporting requirements as
discussed above.
b. Certification Test Groups and Test Vehicle Selection
Manufacturers currently divide their fleet into ``test groups'' for
certification purposes. The test group is EPA's unit of certification;
one certificate is issued per test group. These groupings cover
vehicles with similar emission control system designs expected to have
similar emissions performance.\272\ The factors considered for
determining test groups include combustion cycle, engine type, engine
displacement, number of cylinders and cylinder arrangement, fuel type,
fuel metering system, catalyst construction and precious metal
composition, among others. Vehicles having these features in common are
generally placed in the same test group.\273\ Cars and trucks may be
included in the same test group as long as they have similar emissions
performance (manufacturers frequently produce cars and trucks that have
identical engine designs and emission controls).
---------------------------------------------------------------------------

\272\ 40 CFR 86.1827-01.
\273\ EPA provides for other groupings in certain circumstances,
and can establish its own test groups in cases where the criteria do
not apply. 40 CFR 86.1827-01(b), (c) and (d).
---------------------------------------------------------------------------

EPA recognizes that the Tier 2 test group criteria do not
necessarily relate to CO2 emission levels. For instance,
while some of the criteria, such as combustion cycle, engine type and
displacement, and fuel metering, may have a relationship to
CO2 emissions, others, such as those pertaining to the
catalyst, may not. In fact, there are many vehicle design factors that
affect CO2 generation and emissions but are not included in
EPA's test group criteria.\274\ Most important among these may be
vehicle weight, horsepower, aerodynamics, vehicle size, and performance
features.
---------------------------------------------------------------------------

\274\ EPA noted this potential lack of connection between fuel
economy testing and testing for emissions standard purposes when it
first adopted fuel economy test procedures. See 41 FR at 38677
(Sept. 10, 1976).
---------------------------------------------------------------------------

As described in the proposal, EPA considered but did not propose a
requirement for separate CO2 test groups established around
criteria more directly related to CO2 emissions. Although
CO2-specific test groups might more consistently predict
CO2 emissions of all vehicles in the test group, the
addition of a CO2 test group requirement would greatly
increase the pre-production certification burden for both manufacturers
and EPA. For example, a current Tier 2 test group would need to be
split into two groups if automatic and manual transmissions models had
been included in the same group. Two- and four-wheel drive vehicles in
a current test group would similarly require separation, as would
weight differences among vehicles. This would at least triple the
number of test groups. EPA believes that the added burden of creating
separate CO2 test groups is not warranted or necessary to
maintain an appropriately rigorous certification

[[Page 25472]]

program because the test group data are later replaced by model
specific data which are used as the basis for determining compliance
with a manufacturer's fleet average standard.
For these reasons, EPA will retain the current Tier 2 test group
structure for cars and light trucks in the certification requirements
for CO2. EPA believes that the current test group concept is
also appropriate for N20 and CH4 because the
technologies that are employed to control N2O and
CH4 emissions will generally be the same as those used to
control the criteria pollutants. Vehicle manufacturers agreed with this
assessment and universally supported the use of current Tier 2 test
groups in lieu of developing separate CO2 test groups.
At the time of certification, manufacturers may use the
CO2 emission level from the Tier 2 Emission Data Vehicle as
a surrogate to represent all of the models in the test group. However,
following certification further testing will generally be required for
compliance with the fleet average CO2 standard as described
below. EPA's issuance of a certificate will be conditioned upon the
manufacturer's subsequent model level testing and attainment of the
actual fleet average. Further discussion of these requirements is
presented in Section III.E.6.
As just discussed, the ``worst case'' Emissions Data Vehicle
selected to represent a test group under Tier 2 (40 CFR 86.1828-01) may
not have the highest levels of CO2 in that group. For
instance, there may be a heavier, more powerful configuration that
emits higher CO2, but may, due to the way the catalytic
converter has been matched to the engine, actually have lower
NOX, CO, PM or HC.
Therefore, in lieu of a separate CO2 specific test
group, EPA considered requiring manufacturers to select a
CO2 test vehicle from within the Tier 2 test group that
would be expected, based on good engineering judgment, to have the
highest CO2 emissions within that test group. The
CO2 emissions results from this vehicle would be used to
establish an in-use CO2 emission standard for the test
group. The requirement for a separate, worst case CO2
vehicle would provide EPA with some assurance that all vehicles within
the test group would have CO2 emission levels at or below
those of the selected vehicle, even if there is some variation in the
CO2 control strategies within the test group (such as
different transmission types). Under this approach, the test vehicle
might or might not be the same one that would be selected as worst case
for criteria pollutants. Vehicle manufacturers expressed concern with
this approach as well, and EPA ultimately rejected this approach
because it could have required manufacturers to test two vehicles in
each test group, rather than a single vehicle. This would represent an
added timing burden to manufacturers because they might need to build
additional test vehicles at the time of certification that previously
weren't required to be tested.
Instead, EPA proposed and will adopt provisions that allow a single
Emission Data Vehicle to represent the test group for both Tier 2 and
CO2 certification. The manufacturer will be allowed to
initially apply the Emission Data Vehicle's CO2 emissions
value to all models in the test group, even if other models in the test
group are expected to have higher CO2 emissions. However, as
a condition of the certificate, this surrogate CO2 emissions
value will generally be replaced with actual, model-level
CO2 values based on results from CAFE testing that occurs
later in the model year. This model-level data will become the official
certification test results (as per the conditioned certificate) and
will be used to determine compliance with the fleet average. Only if
the test vehicle is in fact the worst case CO2 vehicle for
the test group could the manufacturer elect to apply the Emission Data
Vehicle emission levels to all models in the test group for purposes of
calculating fleet average emissions. Manufacturers would be unlikely to
make this choice, because doing so would ignore the emissions
performance of vehicle models in their fleet with lower CO2
emissions and would unnecessarily inflate their CO2 fleet
average. Testing at the model level already occurs and data are already
being submitted to EPA for CAFE and labeling purposes, so it would be
an unusual situation that would cause a manufacturer to ignore these
data and choose to accept a higher CO2 fleet average.
Manufacturers will be subject to two standards, the fleet average
standard and the in-use standard for the useful life of the vehicle.
Compliance with the fleet average standard is based on production-
weighted averaging of the test data applied to each model. For each
model, the in-use standard will generally be set at 10% higher than the
level used for that model in calculating the fleet average (see Section
III.E.4).\275\ The certificate will cover both of these standards, and
the manufacturer will have to demonstrate compliance with both of these
standards for purposes of receiving a certificate of conformity. The
certification process for the in-use standard is discussed below in
Section III.E.4.
---------------------------------------------------------------------------

\275\ In cases where configuration or sub-configuration level
data exist, the in-use standard will be set at 10% higher than those
emissions test results. See Section III.E.4.
---------------------------------------------------------------------------

c. Certification Testing Protocols and Procedures
To be consistent with CAFE, EPA will combine the CO2
emissions results from the FTP and HFET tests using the same
calculation method used to determine fuel economy for CAFE purposes.
This approach is appropriate for CO2 because CO2
and fuel economy are so closely related. Other than the fact that fuel
economy is calculated using a harmonic average and CO2
emissions can be calculated using a conventional average, the
calculation methods are very similar. The FTP CO2 data will
be weighted at 55%, and the highway CO2 data at 45%, and
then averaged to determine the combined number. See Section III.B.1 for
more detailed information on CO2 test procedures, Section
III.C.1 on Air Conditioning Emissions, and Section III.B.7 for
N2O and CH4 test procedures.
For the purposes of compliance with the fleet average and in-use
standards, the emissions measured from each test vehicle will include
hydrocarbons (HC) and carbon monoxide (CO), in addition to
CO2. All three of these exhaust constituents are currently
measured and used to determine the amount of fuel burned over a given
test cycle using a ``carbon balance equation'' defined in the
regulations, and thus measurement of these is an integral part of
current fuel economy testing. As explained in Section III.C, it is
important to account for the total carbon content of the fuel.
Therefore the carbon-related combustion products HC and CO must be
included in the calculations along with CO2, and any other
carbon-containing exhaust components such as aldehyde emissions from
alcohol-fueled vehicles. CO emissions are adjusted by a coefficient
that reflects the carbon weight fraction (CWF) of the CO molecule, and
HC emissions are adjusted by a coefficient that reflects the CWF of the
fuel being burned (the molecular weight approach doesn't work since
there are many different hydrocarbon compounds being accounted for).
Thus, EPA will calculate the carbon-related exhaust emissions, also
known as ``CREE,'' of each test vehicle according to the following
formula, where HC, CO, and CO2 are in units of grams per
mile:

[[Page 25473]]

carbon-related exhaust emissions (grams/mile) = CWF*HC + 1.571*CO +
CO2

Where:

CWF = the carbon weight fraction of the test fuel.

As part of the current CAFE and Tier 2 compliance programs, EPA
selects a subset of vehicles for confirmatory testing at its National
Vehicle and Fuel Emissions Laboratory. The purpose of confirmatory
testing is to validate the manufacturer's emissions and/or fuel economy
data. Under this rule, EPA will add CO2, N2O, and
CH4 to the emissions measured in the course of Tier 2 and
CAFE confirmatory testing. The N2O and methane measurement
requirements will begin for model year 2015, when requirements for
manufacturer measurement to comply with the standard also take effect.
The emission values measured at the EPA laboratory will continue to
stand as official, as under existing regulatory programs.
Under current practice, if during EPA's confirmatory fuel economy
testing, the EPA fuel economy value differs from the manufacturer's
value by more than 3%, manufacturers can request a re-test. The re-test
results stand as official, even if they differ by more than 3% from the
manufacturer's value. EPA proposed extending this practice to
CO2 results, but manufacturers commented that this could
lead to duplicative testing and increased test burden. EPA agrees that
the close relationship between CO2 and fuel economy
precludes the need to conduct additional confirmatory tests for both
fuel economy and CO2 to resolve potential discrepancies.
Therefore EPA will continue to allow a re-test request based on a 3% or
greater disparity in manufacturer and EPA confirmatory fuel economy
test values, since a manufacturer's fleet average emissions level would
be established on the basis of model-level testing only (unlike Tier 2
for which a fixed bin standard structure provides the opportunity for a
compliance buffer).
4. Useful Life Compliance
Section 202(a)(1) of the CAA requires emission standards to apply
to vehicles throughout their statutory useful life, as further
described in Section III.A. For emission programs that have fleet
average standards, such as Tier 2 NOX fleet average
standards and the new CO2 standards, the useful life
requirement applies to individual vehicles rather than to the fleet
average standard. For example, in Tier 2 the useful life requirements
apply to the individual emission standard levels or ``bins'' that the
vehicles are certified to, not the fleet average standard. For Tier 2,
the useful life requirement is 10 years \276\ or 120,000 miles with an
optional 15 year or 150,000 mile provision. A similar approach is used
for heavy-duty engines, however a specific Family Emissions Level is
assigned to the engine family at certification, as compared to a pre-
defined bin emissions level as in Tier 2.
---------------------------------------------------------------------------

\276\ 11 years for heavy-light-duty trucks, ref. 40 CFR 86.1805-
12.
---------------------------------------------------------------------------

As noted above, the in-use CO2 standard under the
greenhouse gas program, like Tier 2, will apply to individual vehicles
and is separate from the fleet-average standard. However, unlike the
Tier 2 program and other EPA fleet average standards, the model-level
CO2 test results are themselves used to calculate the fleet
average standard for compliance purposes. This is consistent with the
current CAFE practice, but it means the fleet average standard and the
emission test results used to calculate compliance with the fleet
average standard do not take into account test-to-test variability and
production variability that can affect in-use levels. Since the
CO2 fleet average uses the model level emissions test
results themselves for purposes of calculating the fleet average, EPA
proposed an adjustment factor for the in-use standard to provide some
margin for production and test-to-test variability that could result in
differences between the initial emission test results used to calculate
the fleet average and emission results obtained during subsequent in-
use testing. EPA proposed that each model's in-use CO2
standard would be the model specific level used in calculating the
fleet average, adjusted to be 10% higher.
EPA received significant comment from industry expressing concern
with the in-use standard. The comments focused on concerns about
manufacturer liability for in-use CO2 performance and for
the most part did not address the proposed 10% adjustment level or even
the need for an adjustment to account for variability. Some comments
suggested that an in-use standard is not necessary because in-use
testing is not mandated in the CAA. Others stated that since there is
no evidence that CO2 emission levels increase over time,
there is no need for an in-use standard. Finally, there was a general
concern that failure to meet the in-use standard would result in recall
liability and that recall can only be used in cases where it can be
demonstrated that a ``repair'' can remedy the nonconformity. One
manufacturer provided comments supporting the use of a 10% adjustment
factor for the in-use standard. These comments also recommended that
the 10% adjustment factor be applied to configuration or
subconfiguration data rather than to model-level data unless the lower-
level data were not available. Finally, the manufacturer expressed
concern that a straight 10% adjustment would result in inequity between
high- and low-emitting vehicles.
Section 202(a)(1) specifies that emissions standards are to be
applicable for the useful life of the vehicle. The in-use emissions
standard for CO2 implements this provision. While EPA agrees
that the CAA does not require the Agency to perform in-use testing to
monitor compliance with in-use standards, the Act clearly authorizes
in-use testing. EPA has a long tradition of performing in-use testing
and has found it to be an effective tool in the overall light-duty
vehicle compliance program. EPA continues to believe that it is
appropriate to perform in-use testing and that the evaluation of
individual vehicle performance for all regulated emission constituents,
including CO2, N2O and CH4, is
necessary to ensure compliance with all light-duty requirements. EPA
also believes that the CAA clearly mandates that all emission standards
apply for a vehicle's useful life and that an in-use standard is
therefore necessary.
EPA agrees with industry commenters that there is little evidence
to indicate that CO2 emission levels from current-technology
vehicles increase over time. However, as stated above, the CAA mandates
that all emission standards apply for a vehicle's useful life
regardless of whether the emissions increase over time. In addition,
there are factors other than emission deterioration over time that can
cause in-use emissions to be greater than emission standards. The most
obvious are component defects, production mistakes, and the stacking of
component production and design tolerances. Any one of these can cause
an exceedance of emission standards for individual vehicles or whole
model lines. Finally EPA believes that it is essential to monitor in-
use GHG emissions performance of new technologies, for which there is
currently no in-use experience, as they enter the market. Thus EPA
believes that the value in establishing an in-use standard extends
beyond just addressing emission deterioration over time from current
technology vehicles.
The concern over recall liability in cases where there is no
effective repair remedy has some legitimate basis. For

[[Page 25474]]

example, EPA agrees there would be a concern if a number of vehicles
for a particular model were to have in-use emissions that exceed the
in-use standard, with no effective repair available to remedy the
noncompliance. However, EPA does not anticipate a scenario involving
exceedance of the in-use standard that would cause the Agency to pursue
a recall unless there is a repairable cause of the exceedance. At the
same time, failures to emission-related components, systems, software,
and calibrations do occur that could result in a failure of the in-use
CO2 standard. For example, a defective oxygen sensor that
causes a vehicle to burn excessive fuel could result in higher
CO2 levels that would exceed the in-use standard. While it
is likely that such a problem would affect other emissions as well,
there would still be a demonstratable, repairable problem such that a
recall might be valid. Therefore, EPA believes that a CO2
in-use standard is statutorily required and can serve as a useful tool
for determining compliance with the GHG program.
EPA agrees with the industry comment that it is appropriate where
possible to apply the 10% adjustment factor to the vehicle-level
emission test results, rather than to a model-type value that includes
production weighting factors. If no subconfiguration test data are
available, then the adjustment factor will be applied to the model-type
value. Therefore, EPA is finalizing an in-use standard based on a 10%
multiplicative adjustment factor but the adjustment will be applied to
emissions test results for the vehicle subconfiguration if such data
exist, or to the model-type emissions level used to calculate the fleet
average if subconfiguration test data are not available.
EPA believes that the useful life period established for criteria
pollutants under Tier 2 is also appropriate for CO2. Data
from EPA's current in-use compliance test program indicate that
CO2 emissions from current technology vehicles increase very
little with age and in some cases may actually improve slightly. The
stable CO2 levels are expected because unlike criteria
pollutants, CO2 emissions in current technology vehicles are
not controlled by after treatment systems that may fail with age.
Rather, vehicle CO2 emission levels depend primarily on
fundamental vehicle design characteristics that do not change over
time. Therefore, vehicles designed for a given CO2 emissions
level will be expected to sustain the same emissions profile over their
full useful life.
The CAA requires emission standards to be applicable for the
vehicle's full useful life. Under Tier 2 and other vehicle emission
standard programs, EPA requires manufacturers to demonstrate at the
time of certification that the new vehicles being certified will
continue to meet emission standards throughout their useful life. EPA
allows manufacturers several options for predicting in-use
deterioration, including full vehicle testing, bench-aging specific
components, and application of a deterioration factor based on data
and/or engineering judgment.
In the specific case of CO2, EPA does not currently
anticipate notable deterioration and has therefore determined that an
assigned deterioration factor be applied at the time of certification.
At this time EPA will use an additive assigned deterioration factor of
zero, or a multiplicative factor of one. EPA anticipates that the
deterioration factor will be updated from time to time, as new data
regarding emissions deterioration for CO2 are obtained and
analyzed. Additionally, EPA may consider technology-specific
deterioration factors, should data indicate that certain CO2
control technologies deteriorate differently than others.
During compliance plan discussions prior to the beginning of the
certification process, EPA will explore with each manufacturer any new
technologies that could warrant use of a different deterioration
factor. For any vehicle model determined likely to experience increases
in CO2 emissions over the vehicle's useful life,
manufacturers will not be allowed to use the assigned deterioration
factor but rather will be required to establish an appropriate factor.
If such an instance were to occur, EPA would allow manufacturers to use
the whole-vehicle mileage accumulation method currently offered in
EPA's regulations.\277\
---------------------------------------------------------------------------

\277\ 40 CFR 86.1823-08.
---------------------------------------------------------------------------

N2O and CH4 emissions are directly affected
by vehicle emission control systems. Any of the durability options
offered under EPA's current compliance program can be used to determine
how emissions of N2O and CH4 change over time.
EPA recognizes that manufacturers have not been required to account for
durability effects of N2O and CH4 prior to now.
EPA also realizes that industry will need sufficient time to explore
durability options and become familiar with procedures for determining
deterioration of N2O and CH4. Therefore, until
the 2015 model year, rather than requiring manufacturers to establish a
durability program for N2O and CH4, EPA will
allow manufacturers to attest that vehicles meet the deteriorated, full
useful life standard. If manufacturers choose to comply with the
optional CO2 equivalent standard, EPA will allow the use of
the manufacturer's existing NOX deterioration factor for
N2O and the existing NMOG deterioration factor for
CH4.
a. Ensuring Useful Life Compliance
The CAA requires a vehicle to comply with emission standards over
its regulatory useful life and affords EPA broad authority for the
implementation of this requirement. As such, EPA has authority to
require a manufacturer to remedy any noncompliance issues. The remedy
can range from adjusting a manufacturer's credit balance to the
voluntary or mandatory recall of noncompliant vehicles. These potential
remedies provide manufacturers with a strong incentive to design and
build complying vehicles.
Currently, EPA regulations require manufacturers to conduct in-use
testing as a condition of certification. Specifically, manufacturers
must commit to later procure and test privately-owned vehicles that
have been normally used and maintained. The vehicles are tested to
determine the in-use levels of criteria pollutants when they are in
their first and fourth years of service. This testing is referred to as
the In-Use Verification Program (IUVP) testing, which was first
implemented as part of EPA's CAP 2000 certification program.\278\ The
emissions data collected from IUVP serve several purposes. IUVP results
provide EPA with annual real-world in-use data representing the
majority of certified vehicles. EPA uses IUVP data to identify in-use
problems, validate the accuracy of the certification program, verify
manufacturer durability processes, and support emission modeling
efforts. Manufacturers are required to test low mileage and high
mileage vehicles over the FTP and US06 test cycles. They are also
required to provide evaporative emissions, onboard refueling vapory
recovery (ORVR) emissions and on-board diagnostics (OBD) data.
---------------------------------------------------------------------------

\278\ 64 FR 23906, May 4, 1999.
---------------------------------------------------------------------------

Manufacturers are required to provide data for all regulated
criteria pollutants. Some manufacturers have voluntarily submitted
CO2 data as part of IUVP. EPA proposed that manufacturers
provide CO2, N2O, and CH4 data as part
of the IUVP. EPA also proposed that in order to adequately analyze and
assess

[[Page 25475]]

in-use CO2 results, which are based on the combination of
FTP and highway cycle test results, the highway fuel economy test would
also need to be part of IUVP. The University of California, Santa
Barbara expressed support for including N2O and
CH4 emissions as part of the IUVP. Manufacturer comments
were almost unanimously opposed to including any GHG as part of the
IUVP. Specifically, industry commented that CO2 emissions do
not deteriorate over time and in some cases actually improve. Ford
provided data for several 2004 through 2007 model year vehicles that
indicate CO2 emissions improved an average of 1.42% when
vehicles were tested over 5,000 miles. Manufacturers commented that the
inclusion of a greenhouse gas emissions requirement and the highway
test cycle as part of the IUVP would unnecessarily increase burden on
manufacturers and provide no benefit, since CO2 emissions do
not deteriorate over time. Manufacturers also commented that
N2O and CH4 emissions are very low and by EPA's
own account only represent about 1% of total light-duty vehicle GHG
emissions. They also expressed concern over the cost and burden of
measuring N2O for IUVP, since many manufacturers use
contractor laboratories to assist in their IUVP testing and many of
these facilities do not have the necessary equipment to measure
N2O. They stated that since it was unnecessary to include
CO2 emissions as part of IUVP and since N2O and
CH4 were such small contributors to GHG emissions, it did
not make sense to include N2O and CH4 as part of
the IUVP either. They felt that N2O and CH4 could
be more appropriately handled through attestation or an annual
unregulated emissions report.
As discussed above, although EPA shares the view expressed in
manufacturer comments that historical data demonstrate little
CO2 deterioration, in-use emissions can increase for a
number of reasons other than deterioration over time. For example,
production or design errors can result in increased GHG emissions.
Components that aren't built as they were designed or vehicles
inadvertently assembled improperly or with the wrong parts or with
parts improperly designed can result in GHG emissions greater than
those demonstrated to EPA during the certification process and used in
calculating the manufacturer's fleet average. The ``stacking'' of
component design and production tolerances can also result in in-use
emissions that are greater than those used in calculating a
manufacturer's fleet average.
EPA believes IUVP testing is also important to monitor in-use
versus certification emission levels. Because the emphasis of the GHG
program is on a manufacturer's fleet average standard, it is difficult
for EPA to make an assessment as to whether manufacturer's vehicles are
actually producing the GHG levels claimed in their fleet average
without some in-use data for comparison. For example, EPA has expressed
concern that with the in-use standard based on a 10% adjustment factor,
there would be an incentive for manufacturers to develop their fleet
average utilizing the full range of the 10% in-use standard. The only
way for EPA to assess whether manufacturers are designing and producing
vehicles that meet their respective fleet average standards is for EPA
to be able to review in-use GHG emissions from the IUVP.
Finally EPA does have some concern about potential CO2
emissions deterioration in advanced technologies for which we currently
have no in-use experience or data. Since CAFE has never had an in-use
requirement and today's final regulations are the first ever GHG
standards, there has been no need to focus on GHG emissions in-use as
there will be with the new GHG standards. Many of the advanced
technologies that EPA expects manufacturers to use to meet the GHG
standards have been introduced in production vehicles, but until now
not for the purpose of controlling greenhouse gas emissions. For
example, advanced dual-clutch or seven-speed automatic transmissions,
and start-stop technologies have not been broadly tested in the field
for their long-term CO2 performance. In-use GHG performance
information for vehicles using these technologies is needed for many
reasons, including evaluation of whether allowing use of assigned
deterioration factors for CO2 in lieu of actual
deterioration factors will continue to be appropriate.
Therefore, EPA is finalizing the requirement that all manufacturers
must provide IUVP emissions data for CO2. EPA will also
require manufacturers to perform the highway test cycle as part of
IUVP. Since the CO2 standard reflects a combined value of
FTP and highway results, it is necessary to include the highway
emission test in IUVP to enable EPA to compare an in-use CO2
level with a vehicle's in-use standard. EPA understands that requiring
manufacturers to also measure N2O and CH4 will be
initially challenging, since many manufacturer facilities do not
currently have the proper analytical equipment. To be consistent with
timing of the N2O and CH4 emissions standards for
this rule, N2O and CH4 will not be required for
IUVP until the 2015 model year.
Another component of the CAP 2000 certification program is the In-
Use Confirmatory Program (IUCP). This is a manufacturer-conducted
recall quality in-use test program that can be used as the basis for
EPA to order an emission recall. In order for vehicles tested in the
IUVP to qualify for IUCP, there is a threshold of 1.30 times the
certification emission standard and an additional requirement that at
least 50% of the test vehicles for the test group fail for the same
substance. EPA proposed to exclude IUVP data for CO2,
N2O, and CH4 emissions from the IUCP thresholds.
EPA felt that there was not sufficient data to determine if the
existing IUCP thresholds were appropriate or even applicable to those
emissions. The University of California, Santa Barbara disagreed with
EPA's concerns and recommended that CO2, N2O, and
CH4 emissions all be subject to the IUVP threshold criteria.
Manufacturers commented that since CO2 performance is a
function of vehicle design and cannot be remedied in the field with the
addition or replacement of emissions control devices like traditional
criteria pollutants, it would not be appropriate or necessary to
include IUCP threshold criteria for GHG emissions.
EPA continues to believe that the IUCP is an important part of
EPA's in-use compliance program for traditional criteria pollutants.
For GHG emissions, EPA believes the IUCP will also be a valuable future
tool for achieving compliance. However, there are insufficient data
today to determine whether the current IUCP threshold criteria are
appropriate for GHG emissions. Once EPA can gather more data from the
IUVP program and from EPA's internal surveillance program described
below, EPA will reassess the need to exclude IUCP thresholds, and if
warranted, propose a separate rulemaking establishing IUCP threshold
criteria which may include CO2, N2O, and
CH4 emissions. Therefore, for today's final action, EPA will
exclude IUVP data for CO2, N2O, and
CH4 emissions from the IUCP thresholds.
EPA has also administered its own in-use testing program for light-
duty vehicles under authority of section 207(c) of the CAA for more
than 30 years. In this program, EPA procures and tests representative
privately owned vehicles to determine whether they are complying with
emission standards.

[[Page 25476]]

When testing indicates noncompliance, EPA works with the manufacturer
to determine the cause of the problem and to conduct appropriate
additional testing to determine its extent or the effectiveness of
identified remedies. This program operates in conjunction with the IUVP
program and other sources of information to provide a comprehensive
picture of the compliance profile for the entire fleet and address
compliance problems that are identified. EPA will add CO2,
N2O, and CH4 to the emissions measurements it
collects during surveillance testing.
b. In-Use Compliance Standard
For Tier 2, the in-use standard and the standard used for fleet
average calculation are the same. In-use compliance for an individual
vehicle is determined by comparing the vehicle's in-use emission
results with the emission standard levels or ``bin'' to which the
vehicle is certified rather than to the Tier 2 fleet average standard
for the manufacturer. This is because as part of a fleet average
standard, individual vehicles can be certified to various emission
standard levels, which could be higher or lower than the fleet average
standard. Thus, it would be inappropriate to compare an individual
vehicle to the fleet average, since that vehicle could have been
certified to an emission level that is different than the fleet average
level.
This will also be true for the CO2 fleet average
standard. Therefore, to ensure that an individual vehicle complies with
the CO2 standards in-use, it is necessary to compare the
vehicle's in-use CO2 emission result with the appropriate
model-level certification CO2 level used in determining the
manufacturer's fleet average result.
There is a fundamental difference between the CO2
standards and Tier 2 standards. For Tier 2, the standard level used for
the fleet average calculation is one of eight different emission
levels, or ``bins,'' whereas for the CO2 fleet average
standard, the standard level used for the fleet average calculation is
the model-level certification CO2 result. The Tier 2 fleet
average standard is calculated using the ``bin'' emission level or
standard, not the actual certification emission level of the
certification test vehicle. So no matter how low a manufacturer's
actual certification emission results are, the fleet average is still
calculated based on the ``bin'' level rather than the lower
certification result.\279\ In contrast, the CO2 fleet
average standard will be calculated using the actual vehicle model-
level CO2 values from the certification test vehicles. With
a specified certification emission standard, such as the Tier 2
``bins,'' manufacturers typically attempt to over-comply with the
standard to give themselves some cushion for potentially higher in-use
testing results due to emissions performance deterioration and/or
variability that could result in higher emission levels during
subsequent in-use testing. For our CO2 standards, the
emission level used to calculate the fleet average is the actual
certification vehicle test result, thus manufacturers cannot over
comply since the certification test vehicle result will always be the
value used in determining the CO2 fleet average. If the
manufacturer attempted to design the vehicle to achieve a lower
CO2 value, similar to Tier 2 for in-use purposes, the new
lower CO2 value would simply become the new value used for
calculating the fleet average.
---------------------------------------------------------------------------

\279\ In a similar fashion, the fleet average for heavy-duty
engines is calculated using a Family Emission Level, determined by
the manufacturer, which is different from the emission level of the
test engine.
---------------------------------------------------------------------------

The CO2 fleet average standard is based on the
performance of pre-production technology that is representative of the
point of production, and while there is expected to be limited if any
deterioration in effectiveness for any vehicle during the useful life,
the fleet average standard does not take into account the test-to-test
variability or production variability that can affect in-use levels.
Therefore, EPA believes that unlike Tier 2, it is necessary to have a
different in-use standard for CO2 to account for these
variabilities. EPA proposed an in-use standard that was 10% higher than
the appropriate model-level certification CO2 level used in
determining the manufacturer's fleet average result.
As described above, manufacturers typically design their vehicles
to emit at emission levels considerably below the certification
standards. This intentional difference between the actual emission
level and the emission standard is referred to as ``certification
margin,'' since it is typically the difference between the
certification emission level and the emission standard. The
certification margin can provide manufacturers with some protection
from exceeding emission standards in-use, since the in-use standards
are typically the levels used to calculate the fleet average. For Tier
2, the certification margin is the delta between the specific emission
standard level, or ``bin,'' to which the vehicle is certified, and the
vehicle's certification emission level.
Since the level of the fleet average standard does not reflect this
kind of variability, EPA believes it is appropriate to set an in-use
standard that provides a reasonable cushion for in-use variability that
is beyond a manufacturer's control. EPA proposed a factor of 10% that
would act as a surrogate for a certification margin. The factor would
only be applicable to CO2 emissions, and would be applied to
the model-level test results that are used to establish the model-level
in-use standard.
EPA selected a value of 10% for the in-use standard based on a
review of EPA's fuel economy labeling and CAFE confirmatory test
results for the past several vehicle model years. The EPA data indicate
that it is common for test variability to range between three to six
percent and only on rare occasions to exceed 10%. EPA believes that a
value of 10% should be sufficient to account for testing variability
and any production variability that a manufacturer may encounter. EPA
considered both higher and lower values. The Tier 2 fleet as a whole,
for example, has a certification margin approaching 50%.\280\ However,
there are some fundamental differences between CO2 emissions
and other criteria pollutants in the magnitude of the compounds. Tier 2
NMOG and NOX emission standards are hundredths of a gram per
mile (e.g., 0.07 g/mi NOX & 0.09 g/mi NMOG), whereas the
CO2 standards are four orders of magnitude greater (e.g.,
250 g/mi). Thus EPA does not believe it is appropriate to consider a
value on the order of 50 percent. In addition, little deterioration in
emissions control is expected in-use. The adjustment factor addresses
only one element of what is usually built into a compliance margin.
---------------------------------------------------------------------------

\280\ See pages 39-41 of EPA's Vehicle and Engine Compliance
Activities 2007 Progress Report (EPA-420-R-08-011) published in
October, 2008. This document is available electronically at http://
epa.gov/otaq/about/420r08011.pdf.
---------------------------------------------------------------------------

The intent of the separate in-use standard, based on a 10%
compliance factor adjustment, is to provide a reasonable margin such
that vehicles are not automatically deemed as exceeding standards
simply because of normal variability in test results. EPA has some
concerns however that this in-use compliance factor could be perceived
as providing manufacturers with the ability to design their fleets to
generate CO2 emissions up to 10% higher than the actual
values they use to certify and to calculate the year end fleet average
value that determines compliance with the fleet average standard. This
concern provides additional rationale for

[[Page 25477]]

requiring FTP and HFET IUVP data for CO2 emissions to ensure
that in-use values are not regularly 10% higher than the values used in
the fleet average calculation. If in the course of reviewing a
manufacturer's IUVP data it becomes apparent that a manufacturer's
CO2 results are consistently higher than the values used for
calculation of the fleet average, EPA will discuss the matter with the
manufacturer and consider possible resolutions such as changes to
ensure that the emissions test data more accurately reflect the
emissions level of vehicles at the time of production, increased EPA
confirmatory testing, and other similar measures.
Commenters generally did not comment on whether 10% was the
appropriate level for the adjustment factor. Honda did support use of
the proposed 10% adjustment factor for the in-use standard. But Honda
also recommended that the 10% adjustment factor be applied to
subconfiguration data rather than the model-level data unless there was
no subconfiguration data available. Honda also expressed some concern
over the inequity a straight 10% adjustment would incur between high-
and low-emitting vehicles. They suggested that rather than using an
across-the-board 10% multiplicative adjustment factor applied to the
model-level CO2 value for all vehicles, it would be more
equitable to take the sum of a 5% multiplicative factor applied to the
model-level CO2 value and a 5% factor applied to the
manufacturer's fleet CO2 target.
EPA understands that use of a multiplicative adjustment factor
would result in a higher absolute in-use value for a vehicle that has
higher CO2 than for a vehicle with a lower CO2.
However, this difference is not relevant to the purpose of the
adjustment factor, which is to provide some cushion for test and
production variability. EPA does not believe the difference would be
great enough to confer the higher-emitting vehicles with an unfair
advantage with respect to emissions variability.
Given that the purpose of the in-use standard is to enable a fair
comparison between certification and in-use emission levels, EPA agrees
that it is appropriate to apply the 10% adjustment factor to actual
emission test results rather than to model-type emission levels which
are production weighted. Therefore, EPA is finalizing an in-use
standard that applies a multiplicative 10% adjustment factor to the
subconfiguration emissions values, if such are available. (For
flexible-fuel and dual-fuel vehicles the multiplicative factor will be
applied to the test results on each fuel. In other words, these
vehicles will have two applicable in-use emission standards; one for
operation on the conventional fuel and one for operation on the
alternative fuel.) If no emissions data exist at the subconfiguration
level the adjustment will be applied to the model-type value as
originally proposed. If the in-use emission result for a vehicle
exceeds the emissions level, as applicable, adjusted as just described
by 10%, then the vehicle will have exceeded the in-use emission
standard. The in-use standard will apply to all in-use compliance
testing including IUVP, selective enforcement audits, and EPA's
internal test program.
5. Credit Program Implementation
As described in Section III.E.2 above, for each manufacturer's
model year production, the manufacturer will average the CO2
emissions within each of the two averaging sets (passenger cars and
trucks) and compare that with its respective fleet average standards
(which in turn will have been determined from the appropriate footprint
curve applicable to that model year). In addition to this within-
company averaging, when a manufacturer's fleet average CO2
values of vehicles produced in an averaging set over-complies compared
to the applicable fleet average standard, the manufacturer could
generate credits that it could save for later use (banking) or could
sell or otherwise distribute to another manufacturer (trading). Section
III.C discusses opportunities for manufacturers to improve their fleet
average, beyond the credits that are simply calculated by over-
achieving their applicable fleet average standard. Implementation of
the credit program generally involves two steps: calculation of the
credit amount and reporting the amount and the associated data and
calculations to EPA.
EPA is promulgating two broad types of credit programs under this
rulemaking. One type of credit directly lowers a manufacturer's actual
fleet average by virtue of being applied within the methodology for
calculating the fleet average emissions. Examples of this type of
credit include the credits available for alternative fuel vehicles and
the advanced technology vehicle provisions. The second type of credit
is independent of the calculation of a manufacturer's fleet average.
Rather than giving credit by lowering a manufacturer's fleet average
via a credit mechanism, these credits (in megagrams) are calculated
separately and are simply added to the manufacturer's overall ``bank''
of credits (or debits). Using a fictional example, the remainder of
this section reviews the different types of credits and shows where and
how they are calculated and how they impact a manufacturer's available
credits.
a. Basic Credits: Fleet Average Emissions Are Below the Standard
As just noted, basic credits are earned by a manufacturer's fleet
that performs better than the applicable fleet average standard.
Manufacturers will calculate their fleet average standards (separate
standards are calculated for cars and trucks) using the footprint-based
equations described in Section III.B. A manufacturer's actual end-of-
year fleet average is calculated similarly to the way in which CAFE
values are currently calculated; in fact, the regulations are
essentially identical. The current CAFE calculation methods are in 40
CFR Part 600. As part of this rulemaking, EPA has amended key subparts
and sections of Part 600 to require that fleet average CO2
emissions be calculated in a manner parallel to the way CAFE values are
calculated. First, manufacturers will determine a CO2-
equivalent value for each model type. The CO2-equivalent
value is a summation of the carbon-containing constituents of the
exhaust emissions on a CO2-equivalent basis. For gasoline
and diesel vehicles this simply involves measurement of total
hydrocarbons and carbon monoxide in addition to CO2. The
calculation becomes somewhat more complex for alternative fuel vehicles
due to the different nature of their exhaust emissions. For example,
for ethanol-fueled vehicles, the emission tests must measure ethanol,
methanol, formaldehyde, and acetaldehyde in addition to CO2.
However, all these measurements are currently necessary to determine
fuel economy for the labeling and CAFE programs, and thus no new
testing or data collection will be required.\281\ Second, manufacturers
will calculate a fleet average by weighting the CO2 value
for each model type by the production of that model type, as they
currently do for the CAFE program. Again, this will be done separately
for cars and trucks. Finally, the manufacturer will compare the
calculated standard with the fleet average that is actually achieved to
determine the credits (or debits) that are generated. Both the
determination of the applicable standard and the actual fleet average
will be done after the model

[[Page 25478]]

year is complete and using final model year vehicle production data.
---------------------------------------------------------------------------

\281\ Note that the final rule also provides an option for
manufacturers to incorporate N2O and CH4 in
this calculation at their CO2-equivalent values.
---------------------------------------------------------------------------

Consider a basic hypothetical example where Manufacturer ``A'' has
calculated a car fleet average standard of 300 grams/mile and a car
fleet average of 290 grams/mile (Table III.E.5-1). Further assume that
the manufacturer produced 500,000 cars. The credit is calculated by
taking the difference between the standard and the fleet average (300-
290=10) and multiplying it by the manufacturer's production of 500,000.
This result is then multiplied by the assigned lifetime vehicle miles
travelled (for cars this is 195,264 miles, as discussed in Joint TSD
Chapter 4), then finally divided by 1,000,000 to convert from grams to
total megagrams. The result is the total number of megagrams of credit
generated by the manufacturer's car fleet. The same methodology is used
to calculate the total number of megagrams of deficit, if the
manufacturer was not able to comply with the fleet average standard. In
this example, the result is 976,320 megagrams of credits, as shown in
Table III.E.5-1.

Table III.E.5-1--Summary for Manufacturer A: Earning Basic Credits
----------------------------------------------------------------------------------------------------------------
CO2 Totals
----------------------------------------------------------------------------------------------------------------
Total production................... Conventional: 500,000...... 290 g/mi 500,000
Fleet average standard............. ........................... 300 g/mi
Fleet average...................... ........................... 290 g/mi
Credits............................ [(300-290) x 500,000 x .................. = 954,855 Mg
195,264] / 1,000,000.
----------------------------------------------------------------------------------------------------------------

b. Interim Advanced Technology Vehicle Provisions
The lower exhaust greenhouse gas emissions of some advanced
technology vehicles can directly benefit a manufacturer's fleet
average, thus increasing the amount of fleet average-based credits they
earn (or reducing the amount of debits that would otherwise accrue).
Manufacturers that produce electric vehicles, plug-in hybrid electric
vehicles, or fuel cell electric vehicles will include these vehicles in
the fleet average calculation with their model type emission values. As
described in detail in Section III.C.3, the emissions from electric
vehicles and plug-in hybrid electric vehicles when operating on
electricity will be accounted for by assuming zero emissions (0 g/mi
CO2) for a limited number of vehicles through the 2016 model
year. This interim limited use of 0 g/mi will be allowed for the
technologies specifically noted above and as defined in the
regulations, with the limitation that the vehicles must be certified to
Tier 2 Bin 5 emission standards or cleaner (i.e., advanced technology
vehicles must contribute to criteria pollutant reductions as well as to
greenhouse gas emission reductions).
EPA proposed specific definitions for the vehicle technologies
eligible for these provisions. One manufacturer suggested the following
changes in their comments:
Insert an additional criterion for electric vehicles that
specifically states that an electric vehicle may not have an onboard
combustion engine/generator system.
A minor deletion of text from the definition for ``Fuel
cell.''
The deletion of the requirement that a PHEV have an
equivalent all-electric range of more than 10 miles.
EPA agrees with the first comment. As written in the proposal, a
vehicle with an onboard combustion engine that serves as a generator
would not have been excluded from the definition of electric vehicle.
However, EPA believes it should be. Although such a vehicle might be
propelled by an electric motor directly, if the indirect source of
electricity is an onboard combustion engine then the vehicle is
fundamentally not an electric vehicle. EPA is also adopting the
commenter's proposed rephrasing of the definition for ``Fuel cell,''
which is simpler and clearer. Finally, in the context of the advanced
technology incentive provisions in this final rule, EPA concurs with
the commenter that the requirement that a PHEV have an equivalent all-
electric range of at least ten miles is unnecessary. In the context of
the proposed credit multiplier EPA was concerned that some vehicles
could install a charging system on a limited battery and gain credit
beyond what the limited technology would deserve simply by virtue of
being defined as a PHEV. However, because EPA is not finalizing the
proposed multiplier provisions (see Section III.C.3) and is instead
using as the sole incentive the zero emission tailpipe level as the
compliance value for a manufacturer's fleetwide average, this concern
is no longer valid. Since EPA is not promulgating multipliers, the
concern expressed at proposal no longer applies, and each PHEV will get
a benefit from electricity commensurate with its measured use of grid
electricity, thus EPA is no longer concerned about the multiplier
effect. Thus, EPA is finalizing the following definitions in the
regulations:
Electric vehicle means a motor vehicle that is powered
solely by an electric motor drawing current from a rechargeable energy
storage system, such as from storage batteries or other portable
electrical energy storage devices, including hydrogen fuel cells,
provided that:
[cir] Recharge energy must be drawn from a source off the vehicle,
such as residential electric service;
[cir] The vehicle must be certified to the emission standards of
Bin 1 of Table S04-1 in paragraph (c)(6) of Sec. 86.1811; and
[cir] The vehicle does not have an onboard combustion engine/
generator system as a means of providing electrical energy.
Fuel cell electric vehicle means a motor vehicle propelled
solely by an electric motor where energy for the motor is supplied by a
fuel cell.
Fuel cell means an electrochemical cell that produces
electricity via the non-combustion reaction of a consumable fuel,
typically hydrogen.
Plug-in hybrid electric vehicle (PHEV) means a hybrid
electric vehicle that has the capability to charge the battery from an
off-vehicle electric source, such that the off-vehicle source cannot be
connected to the vehicle while the vehicle is in motion.
With some simplifying assumptions, assume that 25,000 of
Manufacturer A's fleet are now plug-in hybrid electric vehicles with a
calculated CO2 value of 80 g/mi, and the remaining 475,000
are conventional technology vehicles with an average CO2
value of 290 grams/mile. By including the advanced technology PHEVs in
their fleet, Manufacturer A now has more than 2.9 million credits
(Table III.E.5-2).

[[Page 25479]]

Table III.E.5-2--Summary for Manufacturer A: Earning Basic and Interim Advanced Technology Credits
----------------------------------------------------------------------------------------------------------------
CO2 Totals
----------------------------------------------------------------------------------------------------------------
Total production.................. Conventional: 475,000..... 290 g/mi .. 500,000
PHEV: 25,000.............. 80 g/mi
Fleet average standard............ .......................... 300 g/mi
Fleet average..................... [(475,000 x 290) + (25,000 280 g/mi
x 80)] / [500,000].
Credits........................... [(300-280) x 500,000 x ................. = 1,952,640 Mg
195,264] / 1,000,000.
----------------------------------------------------------------------------------------------------------------

c. Flexible-Fuel Vehicle Credits
As noted in Section III.C, treatment of flexible-fuel vehicle (FFV)
credits differs between model years 2012-2015 and 2016 and later. For
the 2012 through 2015 model years the FFV credits will be calculated as
they are in the CAFE program for the same model years, except that
formulae in the final regulations have been modified as needed to do
the calculations in terms of grams per mile of CO2 values
rather than miles per gallon. These credits are integral to the fleet
average calculation and allow the vehicles to be represented by
artificially reduced emissions. To use this credit program, the
CO2 values of FFVs will be represented by the average of two
things: the CO2 value while operating on gasoline and the
CO2 value while operating on the alternative fuel multiplied
by 0.15.
For MY 2012 to 2015 for example, Manufacturer A makes 30,000 FFVs
with CO2 values of 280 g/mi using gasoline and 260 g/mi
using E85. The CO2 value that would represent the FFVs in
the fleet average calculation would be calculated as follows:

FFV emissions = [280 + (260 x 0.15)] / 2 = 160 g/mi

Including these FFVs with the applicable credit in Manufacturer A's
fleet average, as shown below in Table III.E.5-3, further reduces the
fleet average to 256 grams/mile and increases the manufacturer's
credits to about 4.2 million megagrams.

Table III.E.5-3 Summary for Manufacturer A: Earning Basic, Interim Advanced Technology, and Flexible Fuel
Vehicle Credits
----------------------------------------------------------------------------------------------------------------
CO2 Totals
----------------------------------------------------------------------------------------------------------------
Total production.................. Conventional: 445,000..... 290 g/mi .. 500,000
PHEV: 25,000.............. 80 g/mi
FFV: 30,000............... 160 g/mi
Fleet average standard............ .......................... 300 g/mi
Fleet average..................... [(445,000 x 290) + (25,000 272 g/mi
x 80) + 30,000 x 160] /
[500,000].
Credits........................... [(300 - 272) x 500,000 x ................. = 2,733,696 Mg
195,264] / 1,000,000.
----------------------------------------------------------------------------------------------------------------

In the 2016 and later model years, the calculation of FFV emissions
differ substantially from prior years in that the determination of the
CO2 value to represent an FFV model type will be based upon
the actual use of the alternative fuel and on actual emissions while
operating on that fuel. EPA's default assumption in the regulations is
that the alternative fuel is used negligibly, and the CO2
value that will apply to an FFV by default would be the value
determined for operation on conventional fuel. However, if the
manufacturer believes that the alternative fuel is used in real-world
driving and that accounting for this use could improve the fleet
average, the manufacturer has two options. First, the regulations allow
a manufacturer to request that EPA determine an appropriate weighting
value for an alternative fuel to reflect the degree of use of that fuel
in FFVs relative to real-world use of the conventional fuel. Section
III.C describes how EPA might make this determination. Any value
determined by EPA will be published by EPA, and that weighting value
would be available for all manufacturers to use for that fuel. The
second option allows a manufacturer to determine the degree of
alternative fuel use for their own vehicle(s), using a variety of
potential methods. Both the method and the use of the final results
must be approved by EPA before their use is allowed. In either case,
whether EPA supplies the weighting factors or EPA approves a
manufacturer's alternative fuel weighting factors, the CO2
emissions of an FFV in 2016 and later would be as follows (assuming
non-zero use of the alternative fuel):

(W1 x CO2conv) + (W2 x CO2alt),

Where W1 and W2 are the proportion of miles driven using
conventional fuel and alternative fuel, respectively, CO2conv is the
CO2 value while using conventional fuel, and CO2alt is
the CO2 value while using the alternative fuel. In the
example above, for instance, the default CO2 value for
the fictional FFV described above would be the gasoline value of 280
g/mi, and the resulting fleet average and total credits would be 279
g/mi and 2,050,272 megagrams, respectively. However, if the EPA
determines that real-world ethanol use amounts to 40 percent of
driving, then using the equation above the FFV would be included in
the fleet average calculation with a CO2 value of 272 g/
mi, resulting in an overall fleet average of 278 g/mi and total
credit accumulation of 2,147,904 megagrams.
d. Dedicated Alternative Fuel Vehicle Credits
Like the FFV credit program described above, these credits will be
treated differently in the first years of the program than in the 2016
and later model years. In fact, these credits are essentially identical
to the FFV credits except for two things: (1) There is no need to
average CO2 values for gasoline and alternative fuel, and
(2) in 2016 and later there is no demonstration needed to get a benefit
from the alternative fuel. The CO2 values are essentially
determined the same way they are for FFVs operating on the alternative
fuel. For the 2012 through 2015 model years the CO2 test
results are multiplied by the credit adjustment factor of 0.15, and the
result is production-weighted in the fleet average calculation. For
example, assume that Manufacturer A now produces 20,000 dedicated CNG
vehicles with CO2 emissions of 220 grams/mile, in addition
to the FFVs and PHEVs already included in their fleet (Table III.E.5-
4). Prior to the 2016 model year the CO2 emissions

[[Page 25480]]

representing these CNG vehicles will be 33 grams/mile (220 x 0.15).

Table III.E.5-4--Summary for Manufacturer A: Earning Basic, Advanced Technology, Flexible Fuel Vehicle, and
Dedicated Alternative Fuel Vehicle Credits
----------------------------------------------------------------------------------------------------------------
CO2 Totals
----------------------------------------------------------------------------------------------------------------
Total production.................. Conventional: 425,000.... 290 g/mi ... 500,000
PHEV: 25,000............. 80 g/mi ... .........................
FFV: 30,000.............. 160 g/mi ... .........................
CNG: 20,000.............. 33 g/mi ... .........................
Fleet average standard............ ......................... 300 g/mi ... .........................
Fleet average..................... [(425,000 x 290) + 261 g/mi ... .........................
(25,000 x 80) + (30,000
x 160) + (20,000 x 33)] /
[500,000].
Credits........................... [(300-261) x 500,000 x ................. = 3,807,648 Mg
195,264] / 1,000,000.
----------------------------------------------------------------------------------------------------------------

The calculation for 2016 and later will be the same except the 0.15
credit adjustment factor is removed from the equation, and the CNG
vehicles in this example would simply be production-weighted in the
equation using their actual emissions value of 220 grams/mile instead
of the ``credited'' value of 33 grams/mile.
e. Air Conditioning Leakage Credits
Unlike the credit programs described above, air conditioning-
related credits do not affect the overall calculation of the fleet
average or fleet average standard. Whether a manufacturer generates
zero air conditioning credits or many, the calculated fleet average
remains the same. Air conditioning credits are calculated and added to
any credits (or deficit) that results from the fleet average
calculations shown above. Thus, these credits can increase a
manufacturer's credit balance or offset a deficit, but their
calculation is external to the fleet average calculation. As noted in
Section III.C, manufacturers can generate credits for reducing the
leakage of refrigerant from their air conditioning systems. To do this
the manufacturer will identify an air conditioning system improvement,
indicate that they intend to use the improvement to generate credits,
and then calculate an annual leakage rate (grams/year) for that system
based on the method defined by the regulations. Air conditioning
credits will be determined separately for cars and trucks using the car
and truck-specific equations described in Section III.C.
In order to put these credits on the same basis as the basic and
other credits described above, the air conditioning leakage credits
will need to be calculated separately for cars and trucks. Thus, the
resulting grams per mile credit determined from the appropriate car or
truck equation will be multiplied by the lifetime VMT assigned by EPA
(195,264 for cars; 225,865 for trucks), and then divided by 1,000,000
to get the total megagrams of CO2 credits generated by the
improved air conditioning system. Although the calculations are done
separately for cars and trucks, the total megagrams will be summed and
then added to the overall credit balance maintained by the
manufacturer.
For example, assume that Manufacturer A has improved an air
conditioning system that is installed in 250,000 cars and that the
calculated leakage rate is 12 grams/year. Assume that the manufacturer
has also implemented a new refrigerant with a Global Warming Potential
of 850. In this case the credit per air conditioning unit, rounded to
the nearest gram per mile would be:

[13.8 x [1 - (12/16.6 x 850/1,430)] = 7.9 g/mi.

Total megagrams of credits would then be:

[7.9 x 250,000 x 195,264] / 1,000,000 = 385,646 Mg.

These credits would be added directly to a manufacturer's total
balance; thus in this example Manufacturer A would now have, after
consideration of all the above credits, a total of 4,193,294 megagrams
of credits.
f. Air Conditioning Efficiency Credits
As noted in Section III.C.1.b, manufacturers may earn credits for
improvements in air conditioning efficiency that reduce the impact of
the air conditioning system on fuel consumption. These credits are
similar to the air conditioning leakage credits described above, in
that these credits are determined independently from the manufacturer's
fleet average calculation, and the resulting credits are added to the
manufacturer's overall balance for the respective model year. Like the
air conditioning leakage credits, these credits can increase a
manufacturer's credit balance or offset a deficit, but their
calculation is external to the fleet average calculation.
In order to put these credits on the same basis as the basic and
other credits describe above, the air conditioning efficiency credits
are calculated separately for cars and trucks. Thus, the resulting
grams per mile credit determined in the above equation is multiplied by
the lifetime VMT, and then divided by 1,000,000 to get the total
megagrams of efficiency credits generated by the improved air
conditioning system. Although the calculations are done separately for
cars and trucks, the total megagrams can be summed and then added to
the overall credit balance maintained by the manufacturer.
As described in Section III.C, manufacturers will determine their
credit based on selections from a menu of technologies, each of which
provides a gram per mile credit amount. The credits will be summed for
all the technologies implemented by the manufacturer, but cannot exceed
5.7 grams per mile. Once this is done, the calculation is a
straightforward translation of a gram per mile credit to total car or
truck megagrams, using the same methodology described above. For
example, if Manufacturer A implements enough technologies to get the
maximum 5.7 grams per mile for an air conditioning system that sells
250,000 units in cars, the calculation of total credits would be as
follows:

[5.7 x 250,000 x 195,264] / 1,000,000 = 278,251 Mg.

These credits would be added directly to a manufacturer's total
balance; thus in this example Manufacturer A would now have, after
consideration of all the above credits, a total of 4,471,545 megagrams
of credits.
g. Off-Cycle Technology Credits
As described in Section III.C, these credits will be available for
certain new or innovative technologies that achieve

[[Page 25481]]

real-world CO2 reductions that aren't adequately captured on
the city or highway test cycles used to determine compliance with the
fleet average standards. Like the air conditioning credits, these
credits are independent of the fleet average calculation. Section
III.C.4 describes two options for generating these credits: Either
using EPA's 5-cycle fuel economy labeling methodology, or if that
method fails to capture the CO2-reducing impact of the
technology, the manufacturer could propose and use, with EPA approval,
a different analytical approach to determining the credit amount. Like
the air conditioning credits above, these credits will have to be
determined separately for cars and trucks because of the differing
lifetime mileage assumptions between cars and trucks.
Using the 5-cycle approach is relatively straightforward, and
because the 5-cycle formulae account for nationwide variations in
driving conditions, no additional adjustments to the test results would
be necessary. The manufacturer would simply calculate a 5-cycle
CO2 value with the technology installed and operating and
compare it with a 5-cycle CO2 value determined without the
technology installed and/or operating. Existing regulations describe
how to calculate 5-cycle fuel economy values, and the GHG regulations
contain provisions that describe how to calculate 5-cycle
CO2 values (see 40 CFR 600.114-08). The manufacturer will
have to design a test program that accounts for vehicle differences if
the technology is installed in different vehicle types, and enough data
will have to be collected to address data uncertainty issues.
Manufacturers seeking to generate off-cycle credits based on a 5-cycle
analysis will be required to submit a description of their test program
and the results to EPA for approval.
As noted in Section III.C.4, a manufacturer-developed testing, data
collection, and analysis program will require additional EPA approval
and oversight. EPA received considerable comment from environmental and
public interest organizations suggesting that EPA's decisions about
which technologies merit off-cycle credit should be open and public.
EPA agrees that a public process will help ensure a fair review and
alleviate concerns about potential misuse of the off-cycle credit
flexibility. Therefore EPA intends to seek public comment on
manufacturer proposals for off-cycle credit that do not use the 5-cycle
approach to quantify emission reductions. EPA will consider any
comments it receives in determining whether and how much credit is
appropriate. Manufacturers should submit proposals well in advance of
their desired decision date to allow time for these public and EPA
reviews.
Once the demonstration of the CO2 reduction of an off-
cycle technology is complete, and the resulting value accounts for
variations in driving, climate and other conditions across the country,
the two approaches are treated fundamentally the same way and in a way
that parallels the approach for determining the air conditioning
credits described above. Once a gram per mile value is approved by the
EPA, the manufacturer will determine the total credit value by
multiplying the gram per mile per vehicle credit by the production
volume of vehicles with that technology and approved for use of the
credit. This would then be multiplied by the lifetime vehicle miles for
cars or trucks, whichever applies, and divided by 1,000,000 to obtain
total megagrams of CO2 credits. These credits would then be
added to the manufacturer's total balance for the given model year.
Just like the above air conditioning case, an off-cycle technology that
is demonstrated to achieve an average CO2 reduction of 4.4
grams/mile and that is installed in 175,000 cars would generate credits
as follows:

[4.4 x 175,000 x 195,264] / 1,000,000 = 150,353 Mg.
h. End-of-Year Reporting
In general, implementation of the averaging, banking, and trading
(ABT) program, including the calculation of credits and deficits, will
be accomplished via existing reporting mechanisms. EPA's existing
regulations define how manufacturers calculate fleet average miles per
gallon for CAFE compliance purposes. Today's action modifies these
regulations to also require the parallel calculation of fleet average
CO2 levels for car and light truck compliance categories.
These regulations already require an end-of-year report for each model
year, submitted to EPA, which details the test results and calculations
that determine each manufacturer's CAFE levels. EPA will now require a
similar report that includes fleet average CO2 levels and
related information. That can be integrated with the CAFE report at the
manufacturer's option. In addition to requiring reporting of the actual
fleet average achieved, this end-of-year report will also contain the
calculations and data determining the manufacturer's applicable fleet
average standard for that model year. As under the existing Tier 2
program, the report will be required to contain the fleet average
standard, all values required to calculate the fleet average standard,
the actual fleet average CO2 that was achieved, all values
required to calculate the actual fleet average, the number of credits
generated or debits incurred, all the values required to calculate the
credits or debits, the number of credits bought or sold, and the
resulting balance of credits or debits.
Because of the multitude of credit programs that are available
under the greenhouse gas program, the end-of-year report will be
required to have more data and a more defined and specific structure
than the CAFE end-of-year report does today. Although requiring ``all
the data required'' to calculate a given value should be inclusive, the
report will contain some requirements specific to certain types of
credits. For advanced technology credits that apply to vehicles like
electric vehicles and plug-in hybrid electric vehicles, manufacturers
will be required to identify the number and type of these vehicles and
the effect of these credits on their fleet average. The same will be
true for credits due to flexible-fuel and alternative-fuel vehicles,
although for 2016 and later flexible-fuel credits manufacturers may
also have to provide a demonstration of the actual use of the
alternative fuel in-use and the resulting calculations of
CO2 values for such vehicles. For air conditioning leakage
credits manufacturers will have to include a summary of their use of
such credits that will include which air conditioning systems were
subject to such credits, information regarding the vehicle models which
were equipped with credit-earning air conditioning systems, the
production volume of these air conditioning systems, the leakage score
of each air conditioning system generating credits, and the resulting
calculation of leakage credits. Air conditioning efficiency reporting
will be somewhat more complicated given the phase-in of the efficiency
test procedure, and reporting will have to detail compliance with the
phase-in as well as the test results and the resulting efficiency
credits generated. Similar reporting requirements will also apply to
the variety of possible off-cycle credit options, where manufacturers
will have to report the applicable technology, the amount of credit per
unit, the production volume of the technology, and the total credits
from that technology.
Although it is the final end-of-year report, when final production
numbers are known, that will determine the degree of compliance and the
actual values of any credits being generated by

[[Page 25482]]

manufacturers, EPA will expect manufacturers to be prepared to discuss
their compliance approach and their potential use of the variety of
credit options in pre-certification meetings that EPA routinely has
with manufacturers. In addition, and in conjunction with a pre-model
year report required under the CAFE program, the manufacturer will be
required to submit projections of all of the elements described above,
plus any projected credit trading transactions (described below).
Finally, to the extent that there are any credit transactions, the
manufacturer will have to detail in the end-of-year report
documentation on all credit transactions that the manufacturer has
engaged in. Information for each transaction will include: the name of
the credit provider, the name of the credit recipient, the date the
transfer occurred, the quantity of credits transferred, and the model
year in which the credits were earned. The final report is due to EPA
within 90 days of the end of the model year, or no later than March 31
in the calendar year after the calendar year named for the model year.
For example, the final GHG report for the 2012 model year is due no
later than March 31, 2013. Failure by the manufacturer to submit the
annual report in the specified time period will be considered to be a
violation of section 203(a)(1) of the Clean Air Act.
6. Enforcement
As discussed above in Section III.E.5, manufacturers will report to
EPA their fleet average and fleet average standard for a given model
year (reporting separately for each of the car and truck averaging
sets), the credits or deficits generated in the current year, the
balance of credit balances or deficits (taking into account banked
credits, deficit carry-forward, etc. see Section III.E.5), and whether
they were in compliance with the fleet average standard under the terms
of the regulations. EPA will review the annual reports, figures, and
calculations submitted by the manufacturer to determine any
nonconformance.
Each certificate, required prior to introduction into commerce,
will be conditioned upon the manufacturer attaining the CO2
fleet average standard. If a manufacturer fails to meet this condition
and has not generated or purchased enough credits to cover the fleet
average exceedance following the three year deficit carry-forward
(Section III.B.4, then EPA will review the manufacturer's production
for the model year in which the deficit originated and designate which
vehicles caused the fleet average standard to be exceeded.
EPA proposed that the vehicles that would be identified as
nonconforming would come from the most recent model year, and some
comments pointed out that this was inconsistent with how the NLEV and
Tier 2 programs were structured. EPA agrees with these comments and is
finalizing an enforcement structure that is essentially identical to
the one in place for existing programs. EPA would designate as
nonconforming those vehicles with the highest emission values first,
continuing until a number of vehicles equal to the calculated number of
non-complying vehicles as determined above is reached. Those vehicles
would be considered to be not covered by the certificates of conformity
covering those model types. In a test group where only a portion of
vehicles would be deemed nonconforming, EPA would determine the actual
nonconforming vehicles by counting backwards from the last vehicle
produced in that model type. A manufacturer would be subject to
penalties and injunctive orders on an individual vehicle basis for sale
of vehicles not covered by a certificate. This is the same general
mechanism used for the National LEV and Tier 2 corporate average
standards.
Section 205 of the CAA authorizes EPA to assess penalties of up to
$37,500 per vehicle for violations of the requirements or prohibitions
of this rule.\282\ This section of the CAA provides that the agency
shall take the following penalty factors into consideration in
determining the appropriate penalty for any specific case: the gravity
of the violation, the economic benefit or savings (if any) resulting
from the violation, the size of the violator's business, the violator's
history of compliance with this title, action taken to remedy the
violation, the effect of the penalty on the violator's ability to
continue in business, and such other matters as justice may require.
---------------------------------------------------------------------------

\282\ 42 U.S.C. 7524(a), Civil Monetary Penalty Inflation
Adjustment, 69 FR 7121 (Feb. 13, 2004) and Civil Monetary Penalty
Inflation Adjustment Rule, 73 FR 75340 (Dec. 11, 2008).
---------------------------------------------------------------------------

Manufacturer comments expressed concern about potential enforcement
action for violations of the greenhouse gas standards, and the
circumstances under which EPA would impose penalties. Manufacturers
also suggested that EPA should adopt a penalty structure similar to the
one in place under CAFE.
The CAA specifies different civil penalty provisions for
noncompliance than EPCA does, and EPA cannot therefore adopt the CAFE
penalty structure. However, EPA recognizes that it may be appropriate,
should a manufacturer fail to comply with the NHTSA fuel economy
standards as well as the CO2 standard in a case arising out
of the same facts and circumstances, to take into account the civil
penalties that NHTSA has assessed for violations of the CAFE standards
when determining the appropriate penalty amount for violations of the
CO2 emissions standards. This approach is consistent with
EPA's broad discretion to consider ``such other matters as justice may
require,'' and will allow EPA to exercise its discretion to prevent
injustice and ensure that penalties for violations of the
CO2 rule are assessed in a fair and reasonable manner.
The statutory penalty factor that allows EPA to consider ``such
other matters as justice may require'' vests EPA with broad discretion
to reduce the penalty when other adjustment factors prove insufficient
or inappropriate to achieve justice.\283\ The underlying principle of
this penalty factor is to operate as a safety mechanism when necessary
to prevent injustice.\284\
---------------------------------------------------------------------------

\283\ In re Spang & Co., 6 E.A.D. 226, 249 (EAB 1995).
\284\ B.J. Carney Industries, 7 E.A.D. 171, 232, n. 82 (EAB
1997).
---------------------------------------------------------------------------

In other environmental statutes, Congress has specifically required
EPA to consider penalties assessed by other government agencies where
violations arise from the same set of facts. For instance, section
311(b)(8) of the Clean Water Act, 33 U.S.C. 1321(b)(8) authorizes EPA
to consider any other penalty for the same incident when determining
the appropriate Clean Water Act penalty. Likewise, section 113(e) of
the CAA authorizes EPA to consider ``payment by the violator of
penalties previously assessed for the same violation'' when assessing
penalties for certain violations of Title I of the Act.
7. Prohibited Acts in the CAA
Section 203 of the Clean Air Act describes acts that are prohibited
by law. This section and associated regulations apply equally to the
greenhouse gas standards as to any other regulated emission. Acts that
are prohibited by section 203 of the Clean Air Act include the
introduction into commerce or the sale of a vehicle without a
certificate of conformity, removing or otherwise defeating emission
control equipment, the sale or installation of devices designed to
defeat emission controls, and other actions. EPA proposed to include in
the

[[Page 25483]]

regulations a new section that details these prohibited acts. Prior
regulations, such as the NLEV program, had included such a section, and
although there is no burden associated with the regulations or any
specific need to repeat what is in the Clean Air Act, EPA believes that
including this language in the regulations provides clarity and
improves the ease of use and completeness of the regulations. No
comments were received on the proposal, and EPA is finalizing the
section on prohibited acts (see 40 CFR 86.1854-12).
8. Other Certification Issues
a. Carryover/Carry Across Certification Test Data
EPA's certification program for vehicles allows manufacturers to
carry certification test data over and across certification testing
from one model year to the next, when no significant changes to models
are made. EPA will also apply this policy to CO2,
N2O and CH4 certification test data. A
manufacturer may also be eligible to use carryover and carry across
data to demonstrate CO2 fleet average compliance if they
have done so for CAFE purposes.
b. Compliance Fees
The CAA allows EPA to collect fees to cover the costs of issuing
certificates of conformity for the classes of vehicles and engines
covered by this rule. On May 11, 2004, EPA updated its fees regulation
based on a study of the costs associated with its motor vehicle and
engine compliance program (69 FR 51402). At the time that cost study
was conducted the current rulemaking was not considered.
At this time the extent of any added costs to EPA as a result of
this rule is not known. EPA will assess its compliance testing and
other activities associated with the rule and may amend its fees
regulations in the future to include any warranted new costs.
c. Small Entity Exemption
EPA is exempting small entities, and these entities (necessarily)
would not be subject to the certification requirements of this rule.
As discussed in Section III.B.8, businesses meeting the Small
Business Administration (SBA) criterion of a small business as
described in 13 CFR 121.201 would not be subject to the GHG
requirements, pending future regulatory action. EPA proposed that such
entities instead be required to submit a declaration to EPA containing
a detailed written description of how that manufacturer qualifies as a
small entity under the provisions of 13 CFR 121.201. EPA has
reconsidered the need for this additional submission under the
regulations and is deleting it as not necessary. We already have
information on the limited number of small entities that we expect
would receive the benefits of the exemption, and do not need the
proposed regulatory requirement to be able to effectively implement
this exemption for those parties who in fact meet its terms. Small
entities are currently covered by a number of EPA motor vehicle
emission regulations, and they routinely submit information and data on
an annual basis as part of their compliance responsibilities.
As discussed in detail in Section III.B.6, small volume
manufacturers with annual sales volumes of less than 5,000 vehicles
will also be deferred from the CO2 standards, pending future
regulatory action. These manufacturers would still be required to meet
N2O and CH4 standards, however. To qualify for
CO2 standard deferral, manufacturers would need to submit a
declaration to EPA, and would also be required to demonstrate due
diligence in having attempted to first secure credits from other
manufacturers. This declaration would have to be signed by a chief
officer of the company, and would have to be made at least 30 days
prior to the introduction into commerce of any vehicles for each model
year for which the small volume manufacturer status is requested, but
not later than December of the calendar year prior to the model year
for which deferral is requested. For example, if a manufacturer will be
introducing model year 2012 vehicles in October of 2011, then the small
volume manufacturer declaration would be due in September, 2011. If
2012 model year vehicles are not planned for introduction until March,
2012, then the declaration would have to be submitted in December,
2011. Such manufacturers are not automatically exempted from other EPA
regulations for light-duty vehicles and light-duty trucks; therefore,
absent this annual declaration EPA would assume that each manufacturer
was not deferred from compliance with the greenhouse gas standards.
d. Onboard Diagnostics (OBD) and CO2 Regulations
The light-duty on-board diagnostics (OBD) regulations require
manufacturers to detect and identify malfunctions in all monitored
emission-related powertrain systems or components.\285\ Specifically,
the OBD system is required to monitor catalysts, oxygen sensors, engine
misfire, evaporative system leaks, and any other emission control
systems directly intended to control emissions, such as exhaust gas
recirculation (EGR), secondary air, and fuel control systems. The
monitoring threshold for all of these systems or components is 1.5
times the applicable standards, which typically include NMHC, CO,
NOX, and PM. EPA did not propose that CO2
emissions would become one of the applicable standards required to be
monitored by the OBD system. EPA did not propose CO2 become
an applicable standard for OBD because it was confident that many of
the emission-related systems and components currently monitored would
effectively catch any malfunctions related to CO2 emissions.
For example, malfunctions resulting from engine misfire, oxygen
sensors, the EGR system, the secondary air system, and the fuel control
system would all have an impact on CO2 emissions. Thus,
repairs made to any of these systems or components should also result
in an improvement in CO2 emissions. In addition, EPA did not
have data on the feasibility or effectiveness of monitoring various
emission systems and components for CO2 emissions and did
not believe that it would be prudent to include CO2
emissions without such information.
---------------------------------------------------------------------------

\285\ 40 CFR 86.1806-04.
---------------------------------------------------------------------------

EPA did not address whether N2O or CH4
emissions should become applicable standards for OBD monitoring in the
proposal. Several manufacturers felt that EPA's silence on this issue
implied that EPA was proposing that N2O and CH4
emissions become applicable OBD standards. They commented that EPA
should not include them as part of OBD. They felt that adding
N2O and CH4 would significantly increase OBD
development burden, without significant benefit, since any malfunctions
that increase N2O and CH4 would likely be caught
by current OBD system designs. EPA agrees with the manufacturer's
comments on including N2O and CH4 as applicable
standards. Therefore, at this time, EPA is not requiring
CO2, N2O, and CH4 emissions as one of
the applicable standards required for the OBD monitoring threshold. EPA
plans to evaluate OBD monitoring technology, with regard to monitoring
these GHG emissions-related systems and components, and may choose to
propose to include CO2, N2O, and CH4
emissions as part of the OBD requirements in a future regulatory
action.

[[Page 25484]]

e. Applicability of Current High Altitude Provisions to Greenhouse
Gases
Vehicles covered by this rule must meet the CO2,
N2O and CH4 standard at altitude. The CAA
requires emission standards under section 202 for light-duty vehicles
and trucks to apply at all altitudes.\286\ EPA does not expect vehicle
CO2, CH4, or N2O emissions to be
significantly different at high altitudes based on vehicle calibrations
commonly used at all altitudes. Therefore, EPA will retain its current
high altitude regulations so manufacturers will not normally be
required to submit vehicle CO2 test data for high altitude.
Instead, they must submit an engineering evaluation indicating that
common calibration approaches will be utilized at high altitude. Any
deviation in emission control practices employed only at altitude will
need to be included in the auxiliary emission control device (AECD)
descriptions submitted by manufacturers at certification. In addition,
any AECD specific to high altitude will be required to include
emissions data to allow EPA evaluate and quantify any emission impact
and validity of the AECD.
---------------------------------------------------------------------------

\286\ See CAA 206(f).
---------------------------------------------------------------------------

f. Applicability of Standards to Aftermarket Conversions
With the exception of the small entity and small volume exemptions,
EPA's emission standards, including greenhouse gas standards, will
continue to apply as stated in the applicability sections of the
relevant regulations. The greenhouse gas standards are being
incorporated into 40 CFR part 86, subpart S, which includes exhaust and
evaporative emission standards for criteria pollutants. Subpart S
includes requirements for new light-duty vehicles, light-duty trucks,
medium-duty passenger vehicles, Otto-cycle complete heavy-duty
vehicles, and some incomplete light-duty trucks. Subpart S is currently
specifically applicable to aftermarket conversion systems, aftermarket
conversion installers, and aftermarket conversion certifiers, as those
terms are defined in 40 CFR 85.502. EPA expects that some aftermarket
conversion companies will qualify for and seek the small entity and/or
small volume exemption, but those that do not qualify will be required
to meet the applicable emission standards, including the greenhouse gas
standards.
g. Geographical Location of Greenhouse Gas Fleet Vehicles
One manufacturer commented that the CAFE sales area location
defined by Department of Transportation regulations is different than
the EPA sales area location defined by the CAA. DOT regulations require
CAFE compliance \287\ in the 50 states, the District of Columbia, and
Puerto Rico. However, EPA emission certification regulations require
emission compliance \288\ in the 50 states, the District of Columbia,
the Puerto Rico, the Virgin Islands, Guam, American Samoa and the
Commonwealth of the Northern Mariana Islands.
---------------------------------------------------------------------------

\287\ DOT regulations at 49 CFR 525.4(a)(5) read ``The term
customs territory of the United States is used as defined in 19
U.S.C. 1202.'' Section 19 U.S.C. 1202 has been replaced by the
Harmonized Tariff Schedule of the United States. The Harmonized
Tariff Schedule reads in part that ``The term `customs territory of
the United States' * * * includes only the States, the District of
Columbia, and Puerto Rico.''
\288\ Section 216 of the Clean Air Act defines the term commerce
to mean ``(A) commerce between any place in any State and any place
outside thereof; and (B) commerce wholly within the District of
Columbia.''
Section 302(d) of the Clean Air Act reads ``The term `State'
means a State, the District of Columbia, the Commonwealth of Puerto
Rico, the Virgin Islands, Guam, and American Samoa and includes the
Commonwealth of the Northern Mariana Islands.'' In addition, 40 CFR
85.1502(14) regarding the importation of motor vehicles and motor
vehicle engines defines the United States to include ``the States,
the District of Columbia, the Commonwealth of Puerto Rico, the
Commonwealth of the Northern Mariana Islands, Guam, American Samoa,
and the U.S. Virgin Islands.''
---------------------------------------------------------------------------

The comment stated that EPA has the discretion under the CAA to
align the sales area location of production vehicles for the greenhouse
gas fleet with the sales area location for the CAFE fleet and
recommended that EPA amend the definitions in 40 CFR 86.1803
accordingly. This would exclude from greenhouse gas requirements
production vehicles that are introduced into commerce in the Virgin
Islands, Guam, American Samoa, and the Commonwealth of the Northern
Mariana.
Although EPA has tried to harmonize greenhouse gas and CAFE
requirements in this rule to the extent possible, EPA believes that the
approach suggested in comment would be contrary to the requirements of
the Act. EPA does not believe that the Agency has discretion under the
CAA to exclude from greenhouse gas requirements production vehicles
introduced into commerce in the Virgin Islands, Guam, American Samoa,
and the Commonwealth of the Northern Mariana Islands. In addition, this
change would introduce an undesirable level of complexity into the
certification process and result in confusion due to vehicles intended
for commerce in separate geographical locations being covered under a
single certificate. For these reasons, EPA will retain the proposed
greenhouse gas production vehicle sales area location as defined in the
CAA.
9. Miscellaneous Revisions to Existing Regulations
a. Revisions and Additions to Definitions
EPA has amended its definitions of ``engine code,'' ``transmission
class,'' and ``transmission configuration'' in its vehicle
certification regulations (part 86) to conform to the definitions for
those terms in its fuel economy regulations (part 600). The exact terms
in part 86 are used for reporting purposes and are not used for any
compliance purpose (e.g., an engine code will not determine which
vehicle is selected for emission testing). However, the terms are used
for this purpose in part 600 (e.g., engine codes, transmission class,
and transmission configurations are all criteria used to determine
which vehicles are to be tested for the purposes of establishing
corporate average fuel economy). Since the same vehicles tested to
determine corporate average fuel economy will also be tested to
determine fleet average CO2, the same definitions will
apply. Thus EPA has amended its part 86 definitions of the above terms
to conform to the definitions in part 600.
Two provisions have been amended to bring EPA's fuel economy
regulations in Part 600 into conformity with the fleet average
CO2 requirement contained in this rulemaking and with
NHTSA's reform truck regulations. First, the definition of
``footprint'' in this rule is also being added to EPA's part 86 and 600
regulations. This definition is based on the definition promulgated by
NHTSA at 49 CFR 523.2. Second, EPA is amending its model year CAFE
reporting regulations to include the footprint information necessary
for EPA to determine the reformed truck standards and the corporate
average fuel economy. This same information is included in this rule
for fleet average CO2 and fuel economy compliance.
b. Addition of Ethanol Fuel Economy Calculation Procedures
EPA has amended part 600 to add calculation procedures for
determining the carbon-related exhaust emissions and calculating the
fuel economy of vehicles operating on ethanol fuel. Manufacturers have
been using these procedures as needed, but the regulatory

[[Page 25485]]

language--which specifies how to determine the fuel economy of
gasoline, diesel, compressed natural gas, and methanol fueled
vehicles--has not previously been updated to specify procedures for
vehicles operating on ethanol. Under today's rule EPA is requiring use
of a carbon balance approach for ethanol-fueled vehicles that is
similar to the way carbon-related exhaust emissions are calculated for
vehicles operating on other fuels for the purpose of determining fuel
economy and for compliance with the fleet average CO2
standards. The carbon balance formula is similar to the one in place
for methanol, except that ethanol and acetaldehyde emissions must also
be measured for ethanol-fueled vehicles. The carbon balance equation
for determining fuel economy is as follows, where CWF is the carbon
weight fraction of the fuel and CWFexHC is the carbon weight
fraction of the exhaust hydrocarbons:

mpg = (CWF x SG x 3781.8)/((CWFexHC x HC) + (0.429 x CO) +
(0.273 x CO2) + (0.375 x CH3OH) + (0.400 x HCHO)
+ (0.521 x C2H5OH) + (0.545 x
C2H4O)).

The equation for determining the total carbon-related exhaust emissions
for compliance with the CO2 fleet average standards is the
following, where CWFexHC is the carbon weight fraction of
the exhaust hydrocarbons:

CO2-eq = (CWFexHC x HC) + (0.429 x CO) + (0.375 x
CH3OH) + (0.400 x HCHO) + (0.521 x
C2H5OH) + (0.545 x C2H4O) +
CO2.
c. Revision of Electric Vehicle Applicability Provisions
In 1980, EPA issued a rule that provided for the inclusion of
electric vehicles in the CAFE program.\289\ EPA now believes that
certain provisions of the regulations should be updated to reflect the
current state of motor vehicle emission and fuel economy regulations.
In particular, EPA believes that the exemption of electric vehicles in
certain cases from fuel economy labeling and CAFE requirements should
be reevaluated and revised.
---------------------------------------------------------------------------

\289\ 45 FR 49256, July 24, 1980.
---------------------------------------------------------------------------

The 1980 rule created an exemption for electric vehicles from fuel
economy labeling in the following cases: (1) If the electric vehicles
are produced by a company that produces only electric vehicles; and (2)
if the electric vehicles are produced by a company that produces fewer
than 10,000 vehicles of all kinds worldwide. EPA believes that this
exemption language is no longer appropriate and is deleting it from the
affected regulations. First, since 1980 many regulatory provisions have
been put in place to address the concerns of small manufacturers and
enable them to comply with fuel economy and emission programs with
reduced burden. EPA believes that all small volume manufacturers should
compete on a fair and level regulatory playing field and that there is
no longer a need to treat small volume electric vehicles any
differently than small volume manufacturers of other types of vehicles.
Current regulations contain streamlined certification procedures for
small companies, and because electric vehicles emit no direct pollution
there is effectively no certification emission testing burden. For
example, the greenhouse gas regulations contain a provision allowing
the exemption of certain small entities. Meeting the requirements for
fuel economy labeling and CAFE will entail a testing, reporting, and
labeling burden, but these burdens are not extraordinary and should be
applied equally to all small volume manufacturers, regardless of the
fuel that moves their vehicles. EPA has been working with existing
electric vehicle manufacturers on fuel economy labeling, and EPA
believes it is important for the consumer to have impartial, accurate,
and useful label information regarding the energy consumption of these
vehicles. Second, EPCA does not provide for an exemption of electric
vehicles from NHTSA's CAFE program, and NHTSA regulations regarding the
applicability of the CAFE program do not provide an exemption for
electric vehicles. Third, the blanket exemption for any manufacturer of
only electric vehicles assumed at the time that these companies would
all be small, but the exemption language inappropriately did not
account for size and would allow large manufacturers to be exempt as
well. Finally, because of growth expected in the electric vehicle
market in the future, EPA believes that the labeling and CAFE
regulations need to be designed to more specifically accommodate
electric vehicles and to require that consumers be provided with
appropriate information regarding these vehicles. For these reasons EPA
has revised 40 CFR Part 600 applicability regulations such that these
electric vehicle exemptions are deleted starting with the 2012 model
year.
d. Miscellaneous Conforming Regulatory Amendments
EPA has made a number of minor amendments to update the regulations
as needed or to ensure that the regulations are consistent with changes
discussed in this preamble. For example, for consistency with the
ethanol fuel economy calculation procedures discussed above, EPA has
amended regulations where necessary to require the collection of
emissions of ethanol and acetaldehyde. Other changes are made to
applicable sections to remove obsolete regulatory requirements such as
phase-ins related to EPA's Tier 2 emission standards program, and still
other changes are made to better accommodate electric vehicles in EPA
emission control regulations. Not all of these minor amendments are
noted in this preamble, thus the reader should carefully evaluate
regulatory text to ensure a complete understanding of the regulatory
changes being promulgated by EPA.
In the process of amending regulations that vary in applicability
by model year, EPA has several approaches that can be taken. The first
option is to amend an existing section of the regulations. For example,
EPA did this in the final regulations with Sec. 86.111-94. In this
case EPA chose to directly amend this section--which applies to 1994
and later model years as indicated by the suffix after the hyphen--but
ensure that the model year of applicability of the amendments (2015 and
later for N2O measurement) is stated clearly in the
regulatory text. A second option is to create a new section with
specific applicability to the 2012 and later model years; i.e., a
section number with a ``12'' following the hyphen. This approach
typically involves pulling forward all the language from an earlier
model year section, then amending as needed (but it could also involve
a wholesale revision and replacement with entirely new language). For
example, EPA took this approach with Sec. 86.1809-12. Although only
paragraphs (d) and (e) contain revisions pertaining to this greenhouse
gas rule, the remainder of the section is ``pulled forward'' from a
prior model year section (in this case, Sec. 86.1809-10) for
completeness. Thus paragraphs (a) through (c) are unchanged relative to
the prior model year section. Readers should therefore be aware that
sections that are indicated as taking effect in the 2012 model year may
differ in only subtle ways from the prior model year section being
superseded. A third approach (not used in this regulation) is to use
the ``Reserved. For guidance see * * *'' technique. For example, in the
Sec. 86.1809-12, rather than bring forward the existing language from
paragraphs (a) through (c), EPA could have simply put a statement in
the regulations

[[Page 25486]]

directing the reader to refer back to Sec. 86.1809-10 for those
requirements. This method has been used in the past, but is not being
used in this regulation.
10. Warranty, Defect Reporting, and Other Emission-Related Components
Provisions
As outlined in the proposal, Section 207(a) of the Clean Air Act
(CAA) requires manufacturers to provide a defect warranty that warrants
a vehicle is designed to comply with emission standards and will be
free from defects that may cause noncompliance over the specified
warranty period which is 2 years/24,000 miles (whichever is first) or,
for major emission control components, 8 years/80,000 miles. The
warranty covers parts which must function properly to assure continued
compliance with emission standards. The proposal explained that under
the greenhouse gas rule, this coverage would include compliance with
the proposed CO2, CH4, and N2O
standards. The proposal did not discuss the CAA Section 207(b)
performance warranty.
EPA proposed to include air conditioning system components under
the CAA section 207(a) emission warranty in cases where manufacturers
use air conditioning leakage and efficiency credits to comply with the
proposed fleet average CO2 standards. The warranty period of
2 years/24,000 miles would apply. EPA requested comments as to whether
any other parts or components should be designated as ``emission
related parts'' and thus subject to warranty and defect reporting
provisions under this rule.
The Alliance of Automobile Manufacturers (Alliance), Toyota and the
State of New Jersey provided comments. The State of New Jersey
supported EPA's proposal to include motor vehicle air conditioning
system components under the emission warranty provisions. Both the
Alliance and Toyota commented that emission warranty requirements are
not appropriate for mobile air conditioners because (1) in-use
performance of the air conditioning system at levels comparable to a
new vehicle is not needed to achieve the emission levels targeted by
EPA and (2) manufacturer general warranties already cover air
conditioning systems and are typically longer than the two-year/24,000
mile proposed emissions warranty period.
Regarding direct emissions (refrigerant leakage), the Alliance and
Toyota commented that warranty requirements are unnecessary for
refrigerants with a global warming potential (GWP) below 150 because
the environmental impact is negligible even if refrigerants are
released from the system. Regarding indirect emissions (fuel consumed
to power the air conditioning system), the Alliance commented that EPA
should not require warranty coverage of the air conditioning system
because in the vast majority of air conditioning failure modes, the
system stops cooling and ceases operation--either because the critical
moving parts stop moving or because the system is switched off--thereby
actually reducing the indirect CO2 emissions.
EPA received no comments regarding (1) other parts or components
which should be designated as ``emission related parts'' subject to
warranty requirements, (2) defect reporting requirements, or (3) other
requirements associated with warranty and defect reporting requirements
(e.g., voluntary emission-related recall reporting requirements,
performance warranty requirements, voluntary aftermarket parts
certification requirements or tampering requirements.
Defect Warranty. EPA's current policy for defect warranty
requirements is provided in Section 207 of the Act. There are currently
no defect warranty regulations. Congress provided under Section 207(a)
and (b) of the CAA that emission-related components shall be covered
under the 207(a) defect warranty and the 207(b) performance warranty
for the warranty period outlined in section 207(i) of the CAA. For
example, section 207(a) reads in part:

``* * * the manufacturer of each new motor vehicle and new motor
vehicle engine shall warrant to the ultimate purchaser and each
subsequent purchaser that such vehicle or engine is (A) designed,
built and equipped so as to conform at the time of sale with
applicable regulations under section 202, and (B) free from defects
in materials and workmanship which cause such vehicle or engine to
fail to conform with applicable regulations for its useful life (as
determined under sec. 202(d)). In the case of vehicles and engines
manufactured in the model year 1995 and thereafter such warranty
shall require that the vehicle or engine is free from any such
defects for the warranty period provided under subsection (i).''

Section 207(i) reads in part:

``(i) Warranty Period.--
(1) In General.--For purposes of subsection (a)(1) and
subsection (b), the warranty period, effective with respect to new
light-duty trucks and new light-duty vehicles and engines,
manufactured in model year 1995 and thereafter, shall be the first 2
years or 24,000 miles of use (whichever first occurs), except as
provided in paragraph (2). For the purposes of subsection (a)(1) and
subsection (b), for other vehicles and engines the warranty period
shall be the period established by the Administrator by regulation
(promulgated prior to the enactment of the Clean Air Act Amendments
of 1990) for such purposes unless the Administrator subsequently
modifies such regulation.
(2) In the case of a specified major emission control component,
the warranty period for new light-duty trucks and new light-duty
vehicles manufactured in the model year 1995 and thereafter for
purposes of subsection (a)(1) and subsection (b) shall be 8 years or
80,000 miles of use (whichever first occurs). As used in this
paragraph, the term `specified major emission control component'
means only a catalytic converter, an electronic emissions control
unit, and an onboard emissions diagnostic device, except that the
Administrator may designate any other pollution control device or
component as a specified major emission control component if--(A)
the device or component was not in general use on vehicles and
engines manufactured prior to the model year 1990; and (B) the
Administrator determines that the retail cost (exclusive of
installation costs) of such device or component exceeds $200 (in
1989 dollars, adjusted for inflation or deflation) as calculated by
the Administrator at the time of such determination * * *''

Thus, the CAA provides the basis of the warranty requirements
contained in today's final rule, which will cover ``emission related
parts'' necessary to provide compliance with CO2,
CH4, and N2O standards. Emission related parts
would include those parts, systems, components and software installed
for the specific purpose of controlling emissions or those components,
systems, or elements of design which must function properly to assure
continued vehicle emission compliance, including compliance with
CO2, CH4, and N2O standards; (similar
to the current definition of ``emission related parts'' provided in 40
CFR 85.2102(14) for performance warranty requirements). For example,
today's action will extend defect warranty requirements to emission-
related components on advanced technology vehicles such as cylinder
deactivation components or batteries used in hybrid-electric vehicles.
Under today's rule, EPA will extend the defect warranty requirement
to emission-related components necessary to meet CO2,
CH4, and N2O standards, including emission-
related components which are used to obtain optional credits for (1)
certification of advanced technology vehicles, (2) credits for
reduction of air conditioning refrigerant leakage, (3) credits for
improving air conditioning system efficiency, (4) credits for off-cycle
CO2 reducing technologies, and (5) optional early credits
for 2009-2011 model year vehicles outlined in the provisions of 40

[[Page 25487]]

CFR 86.1867-12 (which are required to be reported to EPA after the 2011
model year).
Regarding the comments received by the Alliance and Toyota, that
warranty coverage is not needed for air conditioning components, EPA
believes that the Clean Air Act requires warranty coverage on
components used to demonstrate compliance with the emission standards,
including components used in the optional credit programs for reduction
of air conditioning refrigerant leakage and air conditioning efficiency
improvements. EPA does not have the discretion to forgo warranty
requirements by regulation in today's final rule. Thus, the Agency is
adopting defect warranty requirements for air conditioning components
as proposed.
Effective date of Warranty for Components used to Obtain Early
Credits. Regarding the defect warranty for emission-related components
used to obtain optional early credits for 2009-2011 vehicles, the
defect warranty should provide coverage for these components at the
time the early credits report is submitted to EPA (e.g., no later than
90 days after the end of the 2011 model year). For example, the defect
warranty for early credit components does not have to apply
retroactively (before the manufacturer declares the credits to EPA).
The Agency believes this approach is reasonable, because (1)
manufacturer's early credit plans may not be finalized until after
vehicles have been produced; (2) manufacturers will be provided
satisfactory lead time to provide warranty requirements to customers;
and (3) the manufacturer's basic (bumper-to-bumper) warranty for air
conditioning and other early credit components are typically longer
than the two-year/24,000 mile proposed warranty period which will be
applicable to most early credit components.
Performance Warranty. EPA did not propose any changes to the
current performance warranty requirements, because the performance
warranty preconditions outlined in section 207(b) of the CAA have not
been satisfied. For example, section 207(b) of the CAA comes into play
if EPA issues performance warranty short test regulations and
determines that there are inspection facilities available in the field
to determine when vehicles do not comply with greenhouse gas emission
standards. Once EPA issues performance warranty short test regulations,
then the CAA performance warranty provisions require the manufacturer
to pay for emission-related repairs if a vehicle is properly maintained
and used, and fails the short test and is required to repair the
vehicle. Currently the provisions of 85.2207 and 85.2222 provide
performance warranty short test (commonly called an inspection and
maintenance or I/M test). The provisions of 85.2207 and 85.2222 provide
an I/M test procedure and failure criteria based on an inspection of
the onboard diagnostic (OBD) system of the vehicle. The OBD inspection
procedure in 85.2222 is currently used in most areas of the country
where I/M tests are required. For example, a vehicle fails the OBD test
procedure outlined in 85.2222 if the vehicle's MIL is commanded to be
``on'' during the I/M test procedure.
Although most areas of the country which require I/M testing use
the OBD test procedure outlined in 40 CFR 85.2207 and 85.2222, the NPRM
did not propose that the OBD system would be required to monitor
CO2, CH4 or N2O emission performance,
ref 74 FR 49574 and 74 FR 49755. Therefore, the performance warranty
preconditions in 201(b) of the CAA are not currently in effect for
greenhouse gas CO2 emissions. The performance warranty
continues to apply for criteria pollutants but not for greenhouse
emissions.
Defect Reporting and Voluntary Emission-related Recall Reporting
Requirements. EPA did not propose any changes to the current defect
reporting and voluntary emission-related recall reporting requirements
outlined in the provisions of 40 CFR 85.1901-1909. Although EPA
requested comments, we did not receive any comments on defect reporting
and voluntary emission-related recall reporting requirements. Current
regulations require manufacturers to submit a defect report to EPA
whenever an emission-related defect exists in 25 or more in-use
vehicles or engines of the same model year. The defect report is
required to be submitted to EPA within 15 working days of the time the
manufacturer becomes aware of a defect that affects 25 or more
vehicles. Current regulations require manufacturers to submit to EPA
voluntary emission-related recall reports within 15 working days of the
date when owner notification begins.
Similar to the performance warranty requirements outlined above,
the Agency believes that as proposed, defect reporting and voluntary
emission-related recall reporting requirements would apply to emission-
related components necessary to meet CO2, CH4,
and N2O standards for the useful life of the vehicle,
including emission-related components that are used to obtain optional
credits for (1) certification of advanced technology vehicles, (2)
credits for reduction of air conditioning refrigerant leakage, (3)
credits for improving air conditioning system efficiency, and (4)
credits for off-cycle CO2 reducing technologies, and (5)
optional early credits for 2009-2011 model year vehicles outlined in
the provisions of 40 CFR 86.1867-12 (which are required to be reported
to EPA after the 2011 model year). For early credit components, defect
reporting requirements and voluntary emission-related recall reporting
requirements become effective at the time the early credits report is
submitted to EPA (e.g., no later than 90 days after the end of the 2011
model year).
The final rule includes a minor clarification to the provisions of
40 CFR 85.1902 (b) and (d) to clarify that beginning with the 2012
model year, manufacturers are required to report emission-related
defects and voluntary emission recalls to EPA, including emission-
related defects and voluntary emission recalls related to greenhouse
gas emissions (CH4, N2O and CO2).
11. Light Duty Vehicles and Fuel Economy Labeling
American consumers need accurate and meaningful information about
the environmental and fuel economy performance of new light duty
vehicles. EPA believes it is important that the fuel-economy label
affixed to the new vehicles provide consumers with the critical
information they need to make smart purchase decisions, especially in
light of the expected increase in market share of electric and other
advanced technology vehicles. Consumers may need new and different
information than today's vehicle labels provide in order to help them
understand the energy use and associated cost of owning these electric
and advanced technology vehicles.
Therefore, in proposing this greenhouse gas action, EPA sought
comment on issues surrounding consumer vehicle labeling in general, and
labeling of advanced technology vehicles in particular. EPA
specifically asked for input as to whether today's miles per gallon
fuel economy metric provides adequate information to consumers.
EPA received considerable public input in response to the request
for comment in the proposal. Since the greenhouse gas rule was proposed
in September, 2009, EPA has initiated a separate rulemaking to explore
in detail the information displayed on the fuel economy label and the
methodology for deriving that information. The purpose of the vehicle
labeling rulemaking is to ensure that American consumers

[[Page 25488]]

continue to have the most accurate, meaningful, and useful information
available to them when purchasing new vehicles, and that the
information is presented to them in clear and understandable terms.
EPA will consider all vehicle labeling comments received in
response to the greenhouse gas proposal in its development of the new
labeling rule in coming months. We encourage the interested public to
stay engaged and continue to provide input on this issue in the context
of the vehicle labeling rulemaking.

F. How will this final rule reduce GHG emissions and their associated
effects?

This action is an important step towards curbing steady growth of
GHG emissions from cars and light trucks. In the absence of control,
GHG emissions worldwide and in the U.S. are projected to continue
steady growth. Table III.F-1 shows emissions of CO2,
methane, nitrous oxide and air conditioning refrigerants on a
CO2-equivalent basis for calendar years 2010, 2020, 2030,
2040 and 2050. As shown below, U.S. GHGs are estimated to make up
roughly 17 percent of total worldwide emissions in 2010, and the
contribution of direct emissions from cars and light-trucks to this
U.S. share is growing over time, reaching an estimated 19 percent of
U.S. emissions by 2030 in the absence of control. As discussed later in
this section, this steady rise in GHG emissions is associated with
numerous adverse impacts on human health, food and agriculture, air
quality, and water and forestry resources.

Table III.F-1--Reference Case GHG Emissions by Calendar Year
[MMTCO2eq]
----------------------------------------------------------------------------------------------------------------
2010 2020 2030 2040 2050
----------------------------------------------------------------------------------------------------------------
All Sectors (Worldwide) \a\.................... 41,016 48,059 52,870 56,940 60,209
All Sectors (U.S. Only) \a\.................... 7,118 7,390 7,765 8,101 8,379
U.S. Cars/Light Truck Only \b\................. 1,243 1,293 1,449 1,769 2,219
----------------------------------------------------------------------------------------------------------------
\a\ ADAGE model projections, U.S. EPA.\290\
\b\ MOVES2010 (2010), OMEGA Model (2020-50) U.S. EPA. See RIA Chapter 5.3 for modeling details.

---------------------------------------------------------------------------

\290\ U.S. EPA (2009). ``EPA Analysis of the American Clean
Energy and Security Act of 2009: H.R. 2454 in the 111th Congress.''
U.S. Environmental Protection Agency, Washington, DC USA (http://
www.epa.gov/climatechange/economics/economicanalyses.html). ADAGE
model projections of worldwide and U.S. totals include EISA, and are
provided for context.
---------------------------------------------------------------------------

EPA's GHG rule will result in significant reductions as newer,
cleaner vehicles come into the fleet, and the rule is estimated to have
a measurable impact on world global temperatures. As discussed in
Section I, this GHG rule is part of a joint National Program such that
a large majority of the projected benefits would be achieved jointly
with NHTSA's CAFE standards, which are described in detail in Section
IV. EPA estimates the reductions attributable to the GHG program over
time assuming the model year 2016 standards continue indefinitely post-
2016,\291\ compared to a reference scenario in which the 2011 model
year fuel economy standards continue beyond 2011.
---------------------------------------------------------------------------

\291\ This analysis does not include the EISA requirement for 35
MPG through 2020 or California's Pavley 1 GHG standards. The
standards are intended to supersede these requirements, and the
baseline case for comparison are the emissions that would result
without further action above the currently promulgated fuel economy
standards.
---------------------------------------------------------------------------

Using this approach EPA estimates these standards would cut annual
fleetwide car and light truck tailpipe CO2-eq emissions by
21 percent by 2030, when 90 percent of car and light truck miles will
be travelled by vehicles meeting the new standards. Roughly 20 percent
of these reductions are due to ``upstream'' emission reductions from
gasoline extraction, production and distribution processes as a result
of reduced gasoline demand associated with this rule. Some of the
overall emission reductions also come from projected improvements in
the efficiency of vehicle air conditioning systems, which will
substantially reduce direct emissions of HFCs, one of the most potent
greenhouse gases, as well as indirect emissions of tailpipe
CO2 emissions attributable to reduced engine load from air
conditioning. In total, EPA estimates that compared to a baseline of
indefinite 2011 model year standards, net GHG emission reductions from
the program would be 307 million metric tons CO2-equivalent
(MMTCO2eq) annually by 2030, which represents a reduction of
4 percent of total U.S. GHG emissions and 0.6 percent of total
worldwide GHG emissions projected in that year. This estimate accounts
for all upstream fuel production and distribution emission reductions,
vehicle tailpipe emission reductions including air conditioning
benefits, as well as increased vehicle miles travelled (VMT) due to the
``rebound'' effect discussed in Section III.H. EPA estimates this would
be the equivalent of removing approximately 50 million cars and light
trucks from the road in this timeframe.\292\
---------------------------------------------------------------------------

\292\ Estimated using MOVES2010, the average vehicle in the
light duty fleet emitted 5.1 tons of CO2 during calendar
year 2008.
---------------------------------------------------------------------------

EPA projects the total reduction of the program over the full life
of model year 2012-2016 vehicles to be about 960 MMTCO2eq,
with fuel savings of 78 billion gallons (1.8 billion barrels) of
gasoline over the life of these vehicles, assuming that some
manufacturers take advantage of low-cost HFC reduction strategies to
help meet these standards.
The impacts on global mean temperature and global mean sea level
rise resulting from these emission reductions are discussed in Section
III.F.3.
1. Impact on GHG Emissions
This action will reduce GHG emissions emitted directly from
vehicles due to reduced fuel use and more efficient air conditioning
systems. In addition to these ``downstream'' emissions, reducing
CO2 emissions translates directly to reductions in the
emissions associated with the processes involved in getting petroleum
to the pump, including the extraction and transportation of crude oil,
and the production and distribution of finished gasoline (termed
``upstream'' emissions). Reductions from tailpipe GHG standards grow
over time as the fleet turns over to vehicles subject to the standards,
meaning the benefit of the program will continue as long as the oldest
vehicles in the fleet are replaced by newer, lower CO2
emitting vehicles.
EPA is not projecting any reductions in tailpipe CH4 or
N2O emissions as a result of the emission caps set forth in
this rule, which are meant to prevent emission backsliding and to bring
diesel vehicles equipped with advanced technology aftertreatment, and
other advanced technology vehicles such as lean-burn gasoline vehicles,
into

[[Page 25489]]

alignment with current gasoline vehicle emissions.\293\
---------------------------------------------------------------------------

\293\ EPA is adopting a compliance option whereby manufacturers
can comply with a CO2 equivalent standard in lieu of
meeting the CH4 and N2O standards. This should
have no effect on the estimated GHG reductions attributable to the
rule since a condition of meeting that alternative standard is that
the fleetwide CO2 target remains in place.
---------------------------------------------------------------------------

No substantive comments were received on the emissions modeling
methods or on the greenhouse gas inventories presented in the proposal.
These analyses are updated here to include model revisions and more
recent economic analysis, including revised estimates of future vehicle
sales, fuel prices, and vehicle miles traveled. The primary source for
these data is the AEO 2010 preliminary release.\294\ For more details,
please see the TSD and RIA Chapter 5.
---------------------------------------------------------------------------

\294\ Energy Information Administration. Annual Energy Outlook
2010 Early Release. http://www.eia.doe.gov/oiaf/aeo/.
---------------------------------------------------------------------------

As detailed in the RIA, EPA estimated calendar year tailpipe
CO2 reductions based on pre- and post-control CO2
gram per mile levels from EPA's OMEGA model and assumed to continue
indefinitely into the future, coupled with VMT projections derived from
AEO 2010 Early Release. These estimates reflect the real-world
CO2 emissions reductions projected for the entire U.S.
vehicle fleet in a specified calendar year, including the projected
effect of air conditioning credits, the TLAAS program and FFV credits.
EPA also estimated full lifetime reductions for model years 2012-2016
using pre- and post-control CO2 levels projected by the
OMEGA model, coupled with projected vehicle sales and lifetime mileage
estimates. These estimates reflect the real-world CO2
emissions reductions projected for model years 2012 through 2016
vehicles over their entire life.
This rule allows manufacturers to earn credits for improved vehicle
air conditioning efficiency. Since these improvements are relatively
low cost, EPA projects that manufacturers will take advantage of this
flexibility, leading to reductions from emissions associated with
vehicle air conditioning systems. As explained above, these reductions
will come from both direct emissions of air conditioning refrigerant
over the life of the vehicle and tailpipe CO2 emissions
produced by the increased load of the A/C system on the engine. In
particular, EPA estimates that direct emissions of HFCs, one of the
most potent greenhouse gases, would be reduced 50 percent from light-
duty vehicles when the fleet has turned over to more efficient
vehicles. The fuel savings derived from lower tailpipe CO2
would also lead to reductions in upstream emissions. Our estimated
reductions from the A/C credits program are based on our analysis of
how manufacturers are expected to take advantage of this credit
opportunity in complying with the CO2 fleetwide average
tailpipe standards.
Upstream emission reductions associated with the production and
distribution of fuel were estimated using emission factors from DOE's
GREET1.8 model, with some modifications as detailed in Chapter 5 of the
RIA. These estimates include both international and domestic emission
reductions, since reductions in foreign exports of finished gasoline
and/or crude would make up a significant share of the fuel savings
resulting from the GHG standards. Thus, significant portions of the
upstream GHG emission reductions will occur outside of the U.S.; a
breakdown of projected international versus domestic reductions is
included in the RIA.
a. Calendar Year Reductions for Future Years
Table III.F.1-1 shows reductions estimated from these GHG standards
assuming a pre-control case of 2011 MY standards continuing
indefinitely beyond 2011, and a post-control case in which 2016 MY GHG
standards continue indefinitely beyond 2016.\295\ These reductions are
broken down by upstream and downstream components, including air
conditioning improvements, and also account for the offset from a 10
percent VMT ``rebound'' effect as discussed in Section III.H. Including
the reductions from upstream emissions, total reductions are estimated
to reach 307 MMTCO2eq annually by 2030 (a 21 percent
reduction in U.S. car and light truck emissions), and grow to over 500
MMTCO2eq in 2050 as cleaner vehicles continue to come into
the fleet (a 23 percent reduction in U.S. car and light truck
emissions).
---------------------------------------------------------------------------

\295\ Legally, the 2011 CAFE standards only apply to the 2011
model year and no standards apply to future model years. However, we
do not believe that it would be appropriate to assume that no CAFE
standards would apply beyond the 2011 model year when projecting the
impacts of this rule.

Table III.F.1-1--Projected GHG Reductions
[MMTCO2eq per year]
----------------------------------------------------------------------------------------------------------------
Calendar year
---------------------------------------------------------------
2020 2030 2040 2050
----------------------------------------------------------------------------------------------------------------
Net Reduction *................................. 156.4 307.0 401.5 505.9
Net CO2..................................... 139.1 273.3 360.4 458.7
Net other GHG............................... 17.3 33.7 41.1 47.2
Downstream Reduction............................ 125.2 245.7 320.7 403.0
CO2 (excluding A/C)......................... 101.2 199.5 263.2 335.1
A/C--indirect CO2........................... 10.6 20.2 26.5 33.8
A/C--direct HFCs............................ 13.3 26.0 30.9 34.2
CH4 (rebound effect)........................ 0.0 0.0 0.0 0.0
N2O (rebound effect)........................ 0.0 -0.1 -0.1 -0.1
Upstream Reduction.............................. 31.2 61.3 80.8 102.9
CO2......................................... 27.2 53.5 70.6 89.9
CH4......................................... 3.9 7.6 10.0 12.7
N4O......................................... 0.1 0.3 0.3 0.4
Percent reduction relative to U.S. reference 12.1% 21.2% 22.7% 22.8%
(cars + light trucks)..........................
Percent reduction relative to U.S. reference 2.1% 4.0% 5.0% 6.0%
(all sectors)..................................
Percent reduction relative to worldwide 0.3% 0.6% 0.7% 0.8%
reference......................................
----------------------------------------------------------------------------------------------------------------
* Includes impacts of 10% VMT rebound rate presented in Table III.F.1-3.

[[Page 25490]]

b. Lifetime Reductions for 2012-2016 Model Years
EPA also analyzed the emission reductions over the full life of the
2012-2016 model year cars and trucks affected by this program.\296\
These results, including both upstream and downstream GHG
contributions, are presented in Table III.F.1-2, showing lifetime
reductions of about 960 MMTCO2eq, with fuel savings of 78
billion gallons (1.8 billion barrels) of gasoline.
---------------------------------------------------------------------------

\296\ As detailed in the RIA Chapter 5 and TSD Chapter 4, for
this analysis the full life of the vehicle is represented by average
lifetime mileages for cars (195,000 miles) and trucks (226,000
miles) averaged over calendar years 2012 through 2030, a function of
how far vehicles drive per year and scrappage rates.

Table III.F.1-2--Projected Net GHG Reductions
[MMTCO2eq per year]
------------------------------------------------------------------------
Lifetime GHG Lifetime Fuel
Model year reduction (MMT savings (billion
CO2 EQ) gallons)
------------------------------------------------------------------------
2012................................ 88.9 7.3
2013................................ 130.2 10.5
2014................................ 174.2 13.9
2015................................ 244.2 19.5
2016................................ 324.6 26.5
-----------------------------------
Total Program Benefit........... 962.0 77.7
------------------------------------------------------------------------

c. Impacts of VMT Rebound Effect
As noted above and discussed more fully in Section III.H., the
effect of fuel cost on VMT (``rebound'') was accounted for in our
assessment of economic and environmental impacts of this rule. A 10
percent rebound case was used for this analysis, meaning that VMT for
affected model years is modeled as increasing by 10 percent as much as
the increase in fuel economy; i.e., a 10 percent increase in fuel
economy would yield a 1.0 percent increase in VMT. Results are shown in
Table III.F.1-3; using the 10 percent rebound rate results in an
overall emission increase of 25.0 MMTCO2eq annually in 2030
(this increase is accounted for in the reductions presented in Tables
III.F.1-1 and III.F.1-2). Our estimated changes in CH4 or
N2O emissions as a result of these vehicle GHG standards are
attributed solely to this rebound effect.

Table III.F.1-3--GHG Impact of 10% VMT Rebound a
[MMTCO2eq per year]
----------------------------------------------------------------------------------------------------------------
2020 2030 2040 2050
----------------------------------------------------------------------------------------------------------------
Total GHG Increase.............................. 13.0 25.0 32.9 41.9
Tailpipe & Indirect A/C CO2................. 10.2 19.6 25.8 32.8
Upstream GHGs\ b\........................... 2.8 5.4 7.1 9.1
Tailpipe CH4................................ 0.0 0.0 0.0 0.0
Tailpipe N2O................................ 0.0 0.1 0.1 0.1
----------------------------------------------------------------------------------------------------------------
\a\ These impacts are included in the reductions shown in Table III.F.1-1 and III.F.1-2.
\b\ Upstream rebound impact calculated as upstream total CO2 effect times ratio of downstream tailpipe rebound
CO2 effect to downstream tailpipe total CO2 effect.

d. Analysis of Alternatives
EPA analyzed two alternative scenarios, including 4% and 6% annual
increases in GHG emission standards. In addition to this annual
increase, EPA assumed that manufacturers would use air conditioning
improvements in identical penetrations as in the primary scenario.
Under these assumptions, EPA expects achieved fleetwide average
emission levels of 253 g/mile CO2eq (4%), and 230 g/mile
CO2eq (6%) in 2016.
As in the primary scenario, EPA assumed that the fleet complied
with the standards. For full details on modeling assumptions, please
refer to RIA Chapter 5. EPA's assessment of these alternative
standards, including our response to public comments, is discussed in
Section III.D.

Table III.F.1-4--Calendar Year Impacts of Alternative Scenarios
----------------------------------------------------------------------------------------------------------------
Calendar year
Scenario ---------------------------------------------------
CY 2020 CY 2030 CY 2040 CY 2050
----------------------------------------------------------------------------------------------------------------
Total GHG Reductions (MMT CO2 eq)... Primary............... -156.4 -307.0 -401.5 -505.8
4%.................... -141.9 -286.2 -375.4 -472.9
6%.................... -202.6 -403.4 -529.3 -668.7
Fuel Savings (Billion Gallons Primary............... -12.6 -24.7 -32.6 -41.5
Gasoline Equivalent).
4%.................... -11.3 -22.9 -30.3 -38.6
6%.................... -16.7 -33.2 -43.9 -55.9
----------------------------------------------------------------------------------------------------------------

[[Page 25491]]

Table III.F.1-5--Model Year Impacts of Alternative Scenarios
----------------------------------------------------------------------------------------------------------------
Model year lifetime
Scenario -----------------------------------------------------------------
MY 2012 MY 2013 MY 2014 MY 2015 MY 2016 Total
----------------------------------------------------------------------------------------------------------------
Total GHG Reductions (MMT CO2 Primary........ -88.8 -130.2 -174.2 -244.2 -324.6 -962.0
eq).
4%............. -39.9 -96.6 -155.4 -226.5 -303.6 -822.0
6%............. -61.7 -146.5 -237.0 -332.2 -427.6 -1,204.9
Fuel Savings (Billion Gallons Primary........ -7.3 -10.5 -13.9 -19.5 -26.5 -77.7
Gasoline Equivalent).
4%............. -2.9 -7.1 -12.2 -18.0 -24.6 -64.8
6%............. -4.9 -12.0 -19.4 -27.3 -35.6 -99.1
----------------------------------------------------------------------------------------------------------------

2. Overview of Climate Change Impacts From GHG Emissions
Once emitted, GHGs that are the subject of this regulation can
remain in the atmosphere for decades to centuries, meaning that (1)
their concentrations become well-mixed throughout the global atmosphere
regardless of emission origin, and (2) their effects on climate are
long lasting. GHG emissions come mainly from the combustion of fossil
fuels (coal, oil, and gas), with additional contributions from the
clearing of forests and agricultural activities. The transportation
sector represents a significant portion, 28%, of U.S. GHG
emissions.\297\
---------------------------------------------------------------------------

\297\ U.S. EPA (2009) Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1990-2007. EPA-430-R-09-004, Washington, DC.
---------------------------------------------------------------------------

This section provides a summary of observed and projected changes
in GHG emissions and associated climate change impacts. The source
document for the section below is the Technical Support Document (TSD)
\298\ for EPA's Endangerment and Cause or Contribute Findings Under the
Clean Air Act.\299\ Below is the Executive Summary of the TSD which
provides technical support for the endangerment and cause or contribute
analyses concerning GHG emissions under section 202(a) of the Clean Air
Act. The TSD reviews observed and projected changes in climate based on
current and projected atmospheric GHG concentrations and emissions, as
well as the related impacts and risks from climate change that are
projected in the absence of GHG mitigation actions, including this
action and other U.S. and global actions. The TSD was updated and
revised based on expert technical review and public comment as part of
EPA's rulemaking process for the final Endangerment Findings. The key
findings synthesized here and the information throughout the TSD are
primarily drawn from the assessment reports of the Intergovernmental
Panel on Climate Change (IPCC), the U.S. Climate Change Science Program
(CCSP), the U.S. Global Change Research Program (USGCRP), and the
National Research Council (NRC).\300\
---------------------------------------------------------------------------

\298\ ``Technical Support Document for Endangerment and Cause or
Contribute Findings for Greenhouse Gases Under Section 202(a) of the
Clean Air Act.'' Docket: EPA-HQ-OAR-2009-0472-11292.
\299\ See 74 FR 66496 (Dec. 15, 2009).
\300\ For a complete list of core references from IPCC, USGCRP/
CCSP, NRC and others relied upon for development of the TSD for
EPA's Endangerment and Cause or Contribute Findings see section
1(b), specifically, Table 1.1 of the TSD.
---------------------------------------------------------------------------

a. Observed Trends in Greenhouse Gas Emissions and Concentrations
The primary long-lived GHGs directly emitted by human activities
include carbon dioxide (CO2), methane (CH4),
nitrous oxide (N2O), hydrofluorocarbons (HFCs),
perfluorocarbons (PFCs), and sulfur hexafluoride (SF6).
Greenhouse gases have a warming effect by trapping heat in the
atmosphere that would otherwise escape to space. In 2007, U.S. GHG
emissions were 7,150 teragrams \301\ of CO2 equivalent \302\
(TgCO2eq). The dominant gas emitted is CO2,
mostly from fossil fuel combustion. Methane is the second largest
component of U.S. emissions, followed by N2O and the
fluorinated gases (HFCs, PFCs, and SF6). Electricity
generation is the largest emitting sector (34% of total U.S. GHG
emissions), followed by transportation (28%) and industry (19%).
---------------------------------------------------------------------------

\301\ One teragram (Tg) = 1 million metric tons. 1 metric ton =
1,000 kilograms = 1.102 short tons = 2,205 pounds.
\302\ Long-lived GHGs are compared and summed together on a
CO2-equivalent basis by multiplying each gas by its
global warming potential (GWP), as estimated by IPCC. In accordance
with United Nations Framework Convention on Climate Change (UNFCCC)
reporting procedures, the U.S. quantifies GHG emissions using the
100-year timeframe values for GWPs established in the IPCC Second
Assessment Report.
---------------------------------------------------------------------------

Transportation sources under Section 202(a) \303\ of the Clean Air
Act (passenger cars, light duty trucks, other trucks and buses,
motorcycles, and passenger cooling) emitted 1,649 TgCO2eq in
2007, representing 23% of total U.S. GHG emissions. U.S. transportation
sources under Section 202(a) made up 4.3% of total global GHG emissions
in 2005,\304\ which, in addition to the United States as a whole,
ranked only behind total GHG emissions from China, Russia, and India
but ahead of Japan, Brazil, Germany, and the rest of the world's
countries. In 2005, total U.S. GHG emissions were responsible for 18%
of global emissions, ranking only behind China, which was responsible
for 19% of global GHG emissions. The scope of this action focuses on
GHG emissions under Section 202(a) from passenger cars and light duty
trucks source categories (see Section III.F.1).
---------------------------------------------------------------------------

\303\ Source categories under Section 202(a) of the Clean Air
Act are a subset of source categories considered in the
transportation sector and do not include emissions from non-highway
sources such as boats, rail, aircraft, agricultural equipment,
construction/mining equipment, and other off-road equipment.
\304\ More recent emission data are available for the United
States and other individual countries, but 2005 is the most recent
year for which data for all countries and all gases are available.
---------------------------------------------------------------------------

The global atmospheric CO2 concentration has increased
about 38% from pre-industrial levels to 2009, and almost all of the
increase is due to anthropogenic emissions. The global atmospheric
concentration of CH4 has increased by 149% since pre-
industrial levels (through 2007); and the N2O concentration
has increased by 23% (through 2007). The observed concentration
increase in these gases can also be attributed primarily to
anthropogenic emissions. The industrial fluorinated gases, HFCs, PFCs,
and SF6, have relatively low atmospheric concentrations but
the total radiative forcing due to these gases is increasing rapidly;
these gases are almost entirely anthropogenic in origin.
Historic data show that current atmospheric concentrations of the
two most important directly emitted, long-lived GHGs (CO2
and CH4) are well above the natural range of atmospheric
concentrations compared to at least the last 650,000 years. Atmospheric
GHG concentrations have been increasing because anthropogenic emissions
have been outpacing the rate at which GHGs are removed from the
atmosphere by

[[Page 25492]]

natural processes over timescales of decades to centuries.
b. Observed Effects Associated With Global Elevated Concentrations of
GHGs
Current ambient air concentrations of CO2 and other GHGs
remain well below published exposure thresholds for any direct adverse
health effects, such as respiratory or toxic effects.
The global average net effect of the increase in atmospheric GHG
concentrations, plus other human activities (e.g., land-use change and
aerosol emissions), on the global energy balance since 1750 has been
one of warming. This total net heating effect, referred to as forcing,
is estimated to be +1.6 (+0.6 to +2.4) watts per square meter (W/m\2\),
with much of the range surrounding this estimate due to uncertainties
about the cooling and warming effects of aerosols. However, as aerosol
forcing has more regional variability than the well-mixed, long-lived
GHGs, the global average might not capture some regional effects. The
combined radiative forcing due to the cumulative (i.e., 1750 to 2005)
increase in atmospheric concentrations of CO2,
CH4, and N2O is estimated to be +2.30 (+2.07 to
+2.53) W/m\2\. The rate of increase in positive radiative forcing due
to these three GHGs during the industrial era is very likely to have
been unprecedented in more than 10,000 years.
Warming of the climate system is unequivocal, as is now evident
from observations of increases in global average air and ocean
temperatures, widespread melting of snow and ice, and rising global
average sea level. Global mean surface temperatures have risen by 1.3
0.32 [deg]F (0.74 [deg]C 0.18 [deg]C) over
the last 100 years. Eight of the 10 warmest years on record have
occurred since 2001. Global mean surface temperature was higher during
the last few decades of the 20th century than during any comparable
period during the preceding four centuries.
Most of the observed increase in global average temperatures since
the mid-20th century is very likely due to the observed increase in
anthropogenic GHG concentrations. Climate model simulations suggest
natural forcing alone (i.e., changes in solar irradiance) cannot
explain the observed warming.
U.S. temperatures also warmed during the 20th and into the 21st
century; temperatures are now approximately 1.3 [deg]F (0.7 [deg]C)
warmer than at the start of the 20th century, with an increased rate of
warming over the past 30 years. Both the IPCC \305\ and the CCSP
reports attributed recent North American warming to elevated GHG
concentrations. In the CCSP (2008) report,\306\ the authors find that
for North America, ``more than half of this warming [for the period
1951-2006] is likely the result of human-caused greenhouse gas forcing
of climate change.''
---------------------------------------------------------------------------

\305\ Hegerl, G.C. et al. (2007) Understanding and Attributing
Climate Change. In: Climate Change 2007: The Physical Science Basis.
Contribution of Working Group I to the Fourth Assessment Report of
the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin,
M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, and H.L.
Miller (eds.)]. Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA.
\306\ CCSP (2008) Reanalysis of Historical Climate Data for Key
Atmospheric Features: Implications for Attribution of Causes of
Observed Change. A Report by the U.S. Climate Change Science Program
and the Subcommittee on Global Change Research [Randall Dole, Martin
Hoerling, and Siegfried Schubert (eds.)]. National Oceanic and
Atmospheric Administration, National Climatic Data Center,
Asheville, NC, 156 pp.
---------------------------------------------------------------------------

Observations show that changes are occurring in the amount,
intensity, frequency and type of precipitation. Over the contiguous
United States, total annual precipitation increased by 6.1% from 1901
to 2008. It is likely that there have been increases in the number of
heavy precipitation events within many land regions, even in those
where there has been a reduction in total precipitation amount,
consistent with a warming climate.
There is strong evidence that global sea level gradually rose in
the 20th century and is currently rising at an increased rate. It is
not clear whether the increasing rate of sea level rise is a reflection
of short-term variability or an increase in the longer-term trend.
Nearly all of the Atlantic Ocean shows sea level rise during the last
50 years with the rate of rise reaching a maximum (over 2 millimeters
[mm] per year) in a band along the U.S. east coast running east-
northeast.
Satellite data since 1979 show that annual average Arctic sea ice
extent has shrunk by 4.1% per decade. The size and speed of recent
Arctic summer sea ice loss is highly anomalous relative to the previous
few thousands of years.
Widespread changes in extreme temperatures have been observed in
the last 50 years across all world regions, including the United
States. Cold days, cold nights, and frost have become less frequent,
while hot days, hot nights, and heat waves have become more frequent.
Observational evidence from all continents and most oceans shows
that many natural systems are being affected by regional climate
changes, particularly temperature increases. However, directly
attributing specific regional changes in climate to emissions of GHGs
from human activities is difficult, especially for precipitation.
Ocean CO2 uptake has lowered the average ocean pH
(increased acidity) level by approximately 0.1 since 1750. Consequences
for marine ecosystems can include reduced calcification by shell-
forming organisms, and in the longer term, the dissolution of carbonate
sediments.
Observations show that climate change is currently affecting U.S.
physical and biological systems in significant ways. The consistency of
these observed changes in physical and biological systems and the
observed significant warming likely cannot be explained entirely due to
natural variability or other confounding non-climate factors.
c. Projections of Future Climate Change With Continued Increases in
Elevated GHG Concentrations
Most future scenarios that assume no explicit GHG mitigation
actions (beyond those already enacted) project increasing global GHG
emissions over the century, with climbing GHG concentrations. Carbon
dioxide is expected to remain the dominant anthropogenic GHG over the
course of the 21st century. The radiative forcing associated with the
non-CO2 GHGs is still significant and increasing over time.
Future warming over the course of the 21st century, even under
scenarios of low-emission growth, is very likely to be greater than
observed warming over the past century. According to climate model
simulations summarized by the IPCC,\307\ through about 2030, the global
warming rate is affected little by the choice of different future
emissions scenarios. By the end of the 21st century, projected average
global warming (compared to average temperature around 1990) varies
significantly depending on the emission scenario and climate
sensitivity assumptions, ranging from 3.2 to 7.2 [deg]F (1.8 to 4.0
[deg]C), with an uncertainty range of 2.0 to 11.5 [deg]F (1.1 to 6.4
[deg]C).
---------------------------------------------------------------------------

\307\ Meehl, G.A. et al. (2007) Global Climate Projections. In:
Climate Change 2007: The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M.
Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and
New York, NY, USA.
---------------------------------------------------------------------------

All of the United States is very likely to warm during this
century, and most areas of the United States are expected to warm by
more than the global

[[Page 25493]]

average. The largest warming is projected to occur in winter over
northern parts of Alaska. In western, central and eastern regions of
North America, the projected warming has less seasonal variation and is
not as large, especially near the coast, consistent with less warming
over the oceans.
It is very likely that heat waves will become more intense, more
frequent, and longer lasting in a future warm climate, whereas cold
episodes are projected to decrease significantly.
Increases in the amount of precipitation are very likely in higher
latitudes, while decreases are likely in most subtropical latitudes and
the southwestern United States, continuing observed patterns. The mid-
continental area is expected to experience drying during summer,
indicating a greater risk of drought.
Intensity of precipitation events is projected to increase in the
United States and other regions of the world. More intense
precipitation is expected to increase the risk of flooding and result
in greater runoff and erosion that has the potential for adverse water
quality effects.
It is likely that hurricanes will become more intense, with
stronger peak winds and more heavy precipitation associated with
ongoing increases of tropical sea surface temperatures. Frequency
changes in hurricanes are currently too uncertain for confident
projections.
By the end of the century, global average sea level is projected by
IPCC \308\ to rise between 7.1 and 23 inches (18 and 59 centimeter
[cm]), relative to around 1990, in the absence of increased dynamic ice
sheet loss. Recent rapid changes at the edges of the Greenland and West
Antarctic ice sheets show acceleration of flow and thinning. While an
understanding of these ice sheet processes is incomplete, their
inclusion in models would likely lead to increased sea level
projections for the end of the 21st century.
---------------------------------------------------------------------------

\308\ IPCC (2007) Summary for Policymakers. In: Climate Change
2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M.
Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
---------------------------------------------------------------------------

Sea ice extent is projected to shrink in the Arctic under all IPCC
emissions scenarios.
d. Projected Risks and Impacts Associated With Future Climate Change
Risk to society, ecosystems, and many natural Earth processes
increase with increases in both the rate and magnitude of climate
change. Climate warming may increase the possibility of large, abrupt
regional or global climatic events (e.g., disintegration of the
Greenland Ice Sheet or collapse of the West Antarctic Ice Sheet). The
partial deglaciation of Greenland (and possibly West Antarctica) could
be triggered by a sustained temperature increase of 2 to 7 [deg]F (1 to
4 [deg]C) above 1990 levels. Such warming would cause a 13 to 20 feet
(4 to 6 meter) rise in sea level, which would occur over a time period
of centuries to millennia.
The CCSP \309\ reports that climate change has the potential to
accentuate the disparities already evident in the American health care
system, as many of the expected health effects are likely to fall
disproportionately on the poor, the elderly, the disabled, and the
uninsured. The IPCC \310\ states with very high confidence that climate
change impacts on human health in U.S. cities will be compounded by
population growth and an aging population.
---------------------------------------------------------------------------

\309\ Ebi, K.L., J. Balbus, P.L. Kinney, E. Lipp, D. Mills, M.S.
O'Neill, and M. Wilson (2008) Effects of Global Change on Human
Health. In: Analyses of the effects of global change on human health
and welfare and human systems. A Report by the U.S. Climate Change
Science Program and the Subcommittee on Global Change Research.
[Gamble, J.L. (ed.), K.L. Ebi, F.G. Sussman, T.J. Wilbanks,
(Authors)]. U.S. Environmental Protection Agency, Washington, DC,
USA, pp. 2-1 to 2-78.
\310\ Field, C.B. et al. (2007) North America. In: Climate
Change 2007: Impacts, Adaptation and Vulnerability. Contribution of
Working Group II to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change [M.L. Parry, O.F.
Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and
New York, NY, USA.
---------------------------------------------------------------------------

Severe heat waves are projected to intensify in magnitude and
duration over the portions of the United States where these events
already occur, with potential increases in mortality and morbidity,
especially among the elderly, young, and frail.
Some reduction in the risk of death related to extreme cold is
expected. It is not clear whether reduced mortality from cold will be
greater or less than increased heat-related mortality in the United
States due to climate change.
Increases in regional ozone pollution relative to ozone levels
without climate change are expected due to higher temperatures and
weaker circulation in the United States and other world cities relative
to air quality levels without climate change. Climate change is
expected to increase regional ozone pollution, with associated risks in
respiratory illnesses and premature death. In addition to human health
effects, tropospheric ozone has significant adverse effects on crop
yields, pasture and forest growth, and species composition. The
directional effect of climate change on ambient particulate matter
levels remains uncertain.
Within settlements experiencing climate change, certain parts of
the population may be especially vulnerable; these include the poor,
the elderly, those already in poor health, the disabled, those living
alone, and/or indigenous populations dependent on one or a few
resources. Thus, the potential impacts of climate change raise
environmental justice issues.
The CCSP \311\ concludes that, with increased CO2 and
temperature, the life cycle of grain and oilseed crops will likely
progress more rapidly. But, as temperature rises, these crops will
increasingly begin to experience failure, especially if climate
variability increases and precipitation lessens or becomes more
variable. Furthermore, the marketable yield of many horticultural crops
(e.g., tomatoes, onions, fruits) is very likely to be more sensitive to
climate change than grain and oilseed crops.
---------------------------------------------------------------------------

\311\ Backlund, P., A. Janetos, D.S. Schimel, J. Hatfield, M.G.
Ryan, S.R. Archer, and D. Lettenmaier (2008) Executive Summary. In:
The Effects of Climate Change on Agriculture, Land Resources, Water
Resources, and Biodiversity in the United States. A Report by the
U.S. Climate Change Science Program and the Subcommittee on Global
Change Research. Washington, DC., USA, 362 pp.
---------------------------------------------------------------------------

Higher temperatures will very likely reduce livestock production
during the summer season in some areas, but these losses will very
likely be partially offset by warmer temperatures during the winter
season.
Cold-water fisheries will likely be negatively affected; warm-water
fisheries will generally benefit; and the results for cool-water
fisheries will be mixed, with gains in the northern and losses in the
southern portions of ranges.
Climate change has very likely increased the size and number of
forest fires, insect outbreaks, and tree mortality in the interior
West, the Southwest, and Alaska, and will continue to do so. Over North
America, forest growth and productivity have been observed to increase
since the middle of the 20th century, in part due to observed climate
change. Rising CO2 will very likely increase photosynthesis
for forests, but the increased photosynthesis will likely only increase
wood production in young forests on fertile soils. The combined effects
of expected increased temperature, CO2, nitrogen deposition,
ozone, and forest

[[Page 25494]]

disturbance on soil processes and soil carbon storage remain unclear.
Coastal communities and habitats will be increasingly stressed by
climate change impacts interacting with development and pollution. Sea
level is rising along much of the U.S. coast, and the rate of change
will very likely increase in the future, exacerbating the impacts of
progressive inundation, storm-surge flooding, and shoreline erosion.
Storm impacts are likely to be more severe, especially along the Gulf
and Atlantic coasts. Salt marshes, other coastal habitats, and
dependent species are threatened by sea level rise, fixed structures
blocking landward migration, and changes in vegetation. Population
growth and rising value of infrastructure in coastal areas increases
vulnerability to climate variability and future climate change.
Climate change will likely further constrain already overallocated
water resources in some regions of the United States, increasing
competition among agricultural, municipal, industrial, and ecological
uses. Although water management practices in the United States are
generally advanced, particularly in the West, the reliance on past
conditions as the basis for current and future planning may no longer
be appropriate, as climate change increasingly creates conditions well
outside of historical observations. Rising temperatures will diminish
snowpack and increase evaporation, affecting seasonal availability of
water. In the Great Lakes and major river systems, lower water levels
are likely to exacerbate challenges relating to water quality,
navigation, recreation, hydropower generation, water transfers, and
binational relationships. Decreased water supply and lower water levels
are likely to exacerbate challenges relating to aquatic navigation in
the United States.
Higher water temperatures, increased precipitation intensity, and
longer periods of low flows will exacerbate many forms of water
pollution, potentially making attainment of water quality goals more
difficult. As waters become warmer, the aquatic life they now support
will be replaced by other species better adapted to warmer water. In
the long term, warmer water and changing flow may result in
deterioration of aquatic ecosystems.
Ocean acidification is projected to continue, resulting in the
reduced biological production of marine calcifiers, including corals.
Climate change is likely to affect U.S. energy use and energy
production and physical and institutional infrastructures. It will also
likely interact with and possibly exacerbate ongoing environmental
change and environmental pressures in settlements, particularly in
Alaska where indigenous communities are facing major environmental and
cultural impacts. The U.S. energy sector, which relies heavily on water
for hydropower and cooling capacity, may be adversely impacted by
changes to water supply and quality in reservoirs and other water
bodies. Water infrastructure, including drinking water and wastewater
treatment plants, and sewer and stormwater management systems, will be
at greater risk of flooding, sea level rise and storm surge, low flows,
and other factors that could impair performance.
Disturbances such as wildfires and insect outbreaks are increasing
in the United States and are likely to intensify in a warmer future
with warmer winters, drier soils, and longer growing seasons. Although
recent climate trends have increased vegetation growth, continuing
increases in disturbances are likely to limit carbon storage,
facilitate invasive species, and disrupt ecosystem services.
Over the 21st century, changes in climate will cause species to
shift north and to higher elevations and fundamentally rearrange U.S.
ecosystems. Differential capacities for range shifts and constraints
from development, habitat fragmentation, invasive species, and broken
ecological connections will alter ecosystem structure, function, and
services.
Climate change impacts will vary in nature and magnitude across
different regions of the United States.
Sustained high summer temperatures, heat waves, and
declining air quality are projected in the Northeast,\312\
Southeast,\313\ Southwest,\314\ and Midwest.\315\ Projected climate
change would continue to cause loss of sea ice, glacier retreat,
permafrost thawing, and coastal erosion in Alaska.
---------------------------------------------------------------------------

\312\ Northeast includes West Virginia, Maryland, Delaware,
Pennsylvania, New Jersey, New York, Connecticut, Rhode Island,
Massachusetts, Vermont, New Hampshire, and Maine.
\313\ Southeast includes Kentucky, Virginia, Arkansas,
Tennessee, North Carolina, South Carolina, southeast Texas,
Louisiana, Mississippi, Alabama, Georgia, and Florida.
\314\ Southwest includes California, Nevada, Utah, western
Colorado, Arizona, New Mexico (except the extreme eastern section),
and southwest Texas.
\315\ The Midwest includes Minnesota, Wisconsin, Michigan, Iowa,
Illinois, Indiana, Ohio, and Missouri.
---------------------------------------------------------------------------

Reduced snowpack, earlier spring snowmelt, and increased
likelihood of seasonal summer droughts are projected in the Northeast,
Northwest,\316\ and Alaska. More severe, sustained droughts and water
scarcity are projected in the Southeast, Great Plains,\317\ and
Southwest.
---------------------------------------------------------------------------

\316\ The Northwest includes Washington, Idaho, western Montana,
and Oregon.
\317\ The Great Plains includes central and eastern Montana,
North Dakota, South Dakota, Wyoming, Nebraska, eastern Colorado,
Nebraska, Kansas, extreme eastern New Mexico, central Texas, and
Oklahoma.
---------------------------------------------------------------------------

The Southeast, Midwest, and Northwest in particular are
expected to be impacted by an increased frequency of heavy downpours
and greater flood risk.
Ecosystems of the Southeast, Midwest, Great Plains,
Southwest, Northwest, and Alaska are expected to experience altered
distribution of native species (including local extinctions), more
frequent and intense wildfires, and an increase in insect pest
outbreaks and invasive species.
Sea level rise is expected to increase storm surge height
and strength, flooding, erosion, and wetland loss along the coasts,
particularly in the Northeast, Southeast, and islands.
Warmer water temperatures and ocean acidification are
expected to degrade important aquatic resources of islands and coasts
such as coral reefs and fisheries.
A longer growing season, low levels of warming, and
fertilization effects of carbon dioxide may benefit certain crop
species and forests, particularly in the Northeast and Alaska.
Projected summer rainfall increases in the Pacific islands may augment
limited freshwater supplies. Cold-related mortality is projected to
decrease, especially in the Southeast. In the Midwest in particular,
heating oil demand and snow-related traffic accidents are expected to
decrease.
Climate change impacts in certain regions of the world may
exacerbate problems that raise humanitarian, trade, and national
security issues for the United States. The IPCC \318\ identifies the
most vulnerable world regions as the Arctic, because of the effects of
high rates of projected warming on natural systems; Africa, especially
the sub-Saharan region, because of current low adaptive capacity as
well as climate change; small islands, due to high exposure of
population and infrastructure to risk of sea level rise

[[Page 25495]]

and increased storm surge; and Asian mega-deltas, such as the Ganges-
Brahmaputra and the Zhujiang, due to large populations and high
exposure to sea level rise, storm surge and river flooding. Climate
change has been described as a potential threat multiplier with regard
to national security issues.
---------------------------------------------------------------------------

\318\ Parry, M.L. et al. (2007) Technical Summary. In: Climate
Change 2007: Impacts, Adaptation and Vulnerability. Contribution of
Working Group II to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change [M.L. Parry, O.F.
Canziani, J.P. Palutikof, P.J. van der Linden, and C.E. Hanson
(eds.)], Cambridge University Press, Cambridge, United Kingdom, pp.
23-78.
---------------------------------------------------------------------------

3. Changes in Global Climate Indicators Associated With the Rule's GHG
Emissions Reductions
EPA examined \319\ the reductions in CO2 and other GHGs
associated with this action and analyzed the projected effects on
global mean surface temperature and sea level, two common indicators of
climate change. The analysis projects that this action will reduce
climate warming and sea level rise. Although the projected reductions
are small in overall magnitude by themselves, they are quantifiable and
would contribute to reducing climate change risks. A commenter agreed
that the modeling results showed small, but quantifiable, reductions in
the global atmospheric CO2 concentration, as well as a
reduction in projected global mean surface temperature and sea level
rise, from implementation of this action, across all climate
sensitivities. As such, the commenter encourages the agencies to move
forward with this action while continuing to develop additional, more
stringent vehicle standards beyond 2016.
---------------------------------------------------------------------------

\319\ Using the Model for the Assessment of Greenhouse Gas
Induced Climate Change (MAGICC, http://www.cgd.ucar.edu/cas/wigley/
magicc/), EPA estimated the effects of this action's greenhouse gas
emissions reductions on global mean temperature and sea level.
Please refer to Chapter 7.4 of the RIA for additional information.
---------------------------------------------------------------------------

Another commenter indicated that the projected changes in climate
impacts resulting from this action are small and therefore not
meaningful. EPA disagrees with this view as the reductions may be small
in overall magnitude, but in the global climate change context, they
are quantifiable showing a clear directional signal across a range of
climate sensitivities.^{320 321} EPA therefore determines that
the projected reductions in atmospheric CO2, global mean
temperature and sea level rise are meaningful in the context of this
rule. EPA addresses this point further in the Response to Comments
document. For the final rule, EPA provides an additional climate change
impact analysis for projected changes in ocean pH in the context of
this action. In addition, EPA updated the modeling analysis based on
the revised GHG emission reductions provided in Section III.F.1;
however, the change in modeling results was very small in magnitude.
Based on the reanalysis the results for projected atmospheric
CO2 concentrations are estimated to be reduced by an average
of 2.9 ppm (previously 3.0 ppm), global mean temperature is estimated
to be reduced by 0.006 to 0.015 [deg]C by 2100 (previously 0.007 to
0.016 [deg]C) and sea-level rise is projected to be reduced by
approximately 0.06-0.14cm by 2100 (previously 0.06-0.15cm).
---------------------------------------------------------------------------

\320\ The National Research Council (NRC) 2001 study, Climate
Change Science: An Analysis of Some Key Questions, defines climate
sensitivity as the sensitivity of the climate system to a forcing is
commonly expressed in terms of the global mean temperature change
that would be expected after a time sufficiently long enough for
both the atmosphere and ocean to come to equilibrium with the change
in climate forcing.
\321\ To capture some of the uncertainty in the climate system,
the changes in atmospheric CO2, projected temperatures
and sea level were estimated across the most current
Intergovernmental Panel on Climate Change (IPCC) range of climate
sensitivities, 1.5 [deg]C to 6.0 [deg]C.
---------------------------------------------------------------------------

a. Estimated Projected Reductions in Atmospheric CO2
Concentration, Global Mean Surface Temperatures Sea Level Rise and
Ocean pH
EPA estimated changes in the atmospheric CO2
concentration, global mean surface temperature and sea level to 2100
resulting from the emissions reductions in this action using the Model
for the Assessment of Greenhouse Gas Induced Climate Change (MAGICC,
version 5.3). This widely-used, peer reviewed modeling tool was also
used to project temperature and sea level rise under different
emissions scenarios in the Third and Fourth Assessments of the
Intergovernmental Panel on Climate Change (IPCC).
GHG emissions reductions from Section III.F.1 were applied as net
reductions to a peer reviewed global reference case (or baseline)
emissions scenario to generate an emissions scenario specific to this
action. For the scenario related to this action, all emissions
reductions were assumed to begin in 2012, with zero emissions change in
2011 (from the reference case) followed by emissions linearly
increasing to equal the value supplied in Section III.F.1 for 2020 and
then continuing to 2100. Details about the reference case scenario and
how the emissions reductions were applied to generate the scenario can
be found in the RIA Chapter 7.
Changes in atmospheric CO2 concentration, temperature,
and sea-level for both the reference case and the emissions scenarios
associated with this action were computed using MAGICC. To compute the
reductions in the atmospheric CO2 concentrations as well as
in temperature and sea level resulting from this action, the output
from the scenario associated with this final rule was subtracted from
an existing Global Change Assessment Model (GCAM, formerly MiniCAM)
reference emission scenario. To capture some key uncertainties in the
climate system with the MAGICC model, changes in temperature and sea-
level rise were projected across the most current IPCC range for
climate sensitivities which ranges from 1.5 [deg]C to 6.0 [deg]C
(representing the 90% confidence interval).\322\ This wide range
reflects the uncertainty in this measure of how much the global mean
temperature would rise if the concentration of carbon dioxide in the
atmosphere were to double. Details about this modeling analysis can be
found in the RIA Chapter 7.4.
---------------------------------------------------------------------------

\322\ In IPCC reports, equilibrium climate sensitivity refers to
the equilibrium change in the annual mean global surface temperature
following a doubling of the atmospheric equivalent carbon dioxide
concentration. The IPCC states that climate sensitivity is
``likely'' to be in the range of 2 [deg]C to 4.5 [deg]C, ``very
unlikely'' to be less than 1.5 [deg]C, and ``values substantially
higher than 4.5 [deg]C cannot be excluded.'' IPCC WGI, 2007, Climate
Change 2007--The Physical Science Basis, Contribution of Working
Group I to the Fourth Assessment Report of the IPCC, http://
www.ipcc.ch/.
---------------------------------------------------------------------------

The results of this modeling, summarized in Table III.F.3-1, show
small, but quantifiable, reductions in atmospheric CO2
concentrations, projected global mean surface temperature and sea level
resulting from this action, across all climate sensitivities. As a
result of the emission reductions from this action, the atmospheric
CO2 concentration is projected to be reduced by an average
of 2.9 parts per million (ppm), the global mean temperature is
projected to be reduced by approximately 0.006-0.015[deg]C by 2100, and
global mean sea level rise is projected to be reduced by approximately
0.06-0.14cm by 2100. The reductions are small relative to the IPCC's
2100 ``best estimates'' for global mean temperature increases (1.8-4.0
[deg]C) and sea level rise (0.20-0.59m) for all global GHG emissions
sources for a range of emissions scenarios. EPA used a peer reviewed
model, the MAGICC model, to do this analysis. This analysis is specific
to this rule and therefore does not come from previously published
work. Further discussion of EPA's modeling analysis is found in the
final RIA.

[[Page 25496]]

Table III.F.3-1--Effect of GHG Emissions Reductions on Projected Changes in Global Climate for the Final
Vehicles Rulemaking
[For climate sensitivities ranging from 1.5-6 [deg]C]
----------------------------------------------------------------------------------------------------------------
Measure Units Year Projected change
----------------------------------------------------------------------------------------------------------------
Atmospheric CO2 Concentration................. ppm.............................. 2100 -2.7-3.1
Global Mean Surface Temperature............... [deg]C........................... 2100 -0.006-0.015
Sea Level Rise................................ Cm............................... 2100 -0.06-0.14
Ocean pH...................................... pH units......................... 2100 0.0014
----------------------------------------------------------------------------------------------------------------

As a substantial portion of CO2 emitted into the
atmosphere is not removed by natural processes for millennia, each unit
of CO2 not emitted into the atmosphere avoids essentially
permanent climate change on centennial time scales. Though the
magnitude of the avoided climate change projected here is small, these
reductions would represent a reduction in the adverse risks associated
with climate change (though these risks were not formally estimated for
this action) across all climate sensitivities.
The IPCC \323\ has noted that ocean acidification due to the direct
effects of elevated CO2 concentrations will impair a wide
range of planktonic and other marine organisms that use aragonite to
make their shells or skeletons. EPA used the Program CO2SYS,\324\
version 1.05 to estimate projected changes in tropical ocean pH based
on the atmospheric CO2 concentration reductions resulting
from this action and other specified input conditions (e.g., sea
surface temperature characteristic of tropical waters). The program
performs calculations relating parameters of the carbon dioxide
(CO2) system in seawater. EPA used the program to calculate
ocean pH as a function of atmospheric CO2, among other
specified input conditions. Based on the projected atmospheric
CO2 concentration reductions (average of 2.9 ppm by 2100)
that would result from this rule, the program calculates an increase in
ocean pH of about 0.0014 pH units in 2100. Thus, this analysis
indicates the projected decrease in atmospheric CO2
concentrations from today's rule would result in an increase in ocean
pH.
---------------------------------------------------------------------------

\323\ Fischlin, A. et al. (2007) Ecosystems, their Properties,
Goods, and Services. In: Climate Change 2007: Impacts, Adaptation
and Vulnerability. Contribution of Working Group II to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change
[M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and
C.E. Hanson (eds.)]. Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA.
\324\ Lewis, E., and D. W. R. Wallace. 1998. Program Developed
for CO2 System Calculations. ORNL/CDIAC-105. Carbon
Dioxide Information Analysis Center, Oak Ridge National Laboratory,
U.S. Department of Energy, Oak Ridge, Tennessee.
---------------------------------------------------------------------------

EPA's analysis of the rule's effect on global climate conditions is
intended to quantify these potential reductions using the best
available science. While EPA's modeling results of the effect of this
rule alone show small differences in climate effects (CO2
concentration, temperature, sea-level rise, ocean pH), when expressed
in terms of global climate endpoints and global GHG emissions, they
yield results that are repeatable and consistent within the modeling
frameworks used.

G. How will the standards impact non-GHG emissions and their associated
effects?

In addition to reducing the emissions of greenhouse gases, this
rule will influence the emissions of ``criteria'' air pollutants and
air toxics (i.e., hazardous air pollutants). The criteria air
pollutants include carbon monoxide (CO), fine particulate matter
(PM2.5), sulfur dioxide (SOX) and the ozone precursors
hydrocarbons (VOC) and oxides of nitrogen (NOX); the air
toxics include benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and
acrolein. Our estimates of these non-GHG emission impacts from the GHG
program are shown by pollutant in Table III.G-1 and Table III.G-2 in
total, and broken down by the two drivers of these changes: (a)
``Upstream'' emission reductions due to decreased extraction,
production and distribution of motor gasoline; and (b) ``downstream''
emission increases, reflecting the effects of VMT rebound (discussed in
Sections III.F and III.H) and the effects of our assumptions about
ethanol-blended fuel (E10), as discussed below. Total program impacts
on criteria and toxics emissions are discussed below, followed by
individual discussions of the upstream and downstream impacts. Those
are followed by discussions of the effects on air quality, health, and
other environmental concerns.
As in the proposal, for this analysis we attribute decreased fuel
consumption from this program to gasoline only, while assuming no
effect on volumes of ethanol and other renewable fuels because they are
mandated under the Renewable Fuel Standard (RFS2). However, because
this rule does not assume RFS2 volumes of ethanol in the baseline, the
result is a greater projected market share of E10 in the control
case.\325\ In fact, the GHG standards will not be affecting the market
share of E10, because EPA's analysis for the RFS2 rule predicts 100%
E10 penetration by 2014.\326\
---------------------------------------------------------------------------

\325\ When this rule's analysis was initiated, the RFS2 rule was
not yet final. Therefore, it assumes the ethanol volumes in Annual
Energy Outlook 2007 (U.S. Energy Information Administration, Annual
Energy Outlook 2007, Transportation Demand Sector Supplemental
Table. http://www.eia.doe.gov/oiaf/archive/aeo07/supplement/
index.html)
\326\ EPA 2010, Renewable Fuel Standard Program (RFS2)
Regulatory Impact Analysis. EPA-420-R-10-006. February 2010. Docket
EPA-HQ-OAR-2009-0472-11332. See also 75 FR 14670, March 26, 2010.
---------------------------------------------------------------------------

The amount of E10 affects downstream non-GHG emissions. In the
proposal, EPA stated these same fuel assumptions and qualitatively
noted that there were likely unquantified impacts on non-GHG emissions
between the two cases. In DRIA Chapter 5, EPA indicated its plans to
quantify these impacts in the air quality modeling and in the final
rule inventories. Upstream emission impacts depend only on fuel
volumes, so the impacts presented here reflect only the reduced
gasoline consumption.
The inventories presented in this rulemaking include an analysis of
these fuel effects which was conducted using EPA's Motor Vehicle
Emission Simulator (MOVES2010). The most notable impact, although still
relatively slight, is a 2.2 percent increase in 2030 in national
acetaldehyde emissions over the baseline scenario. It should be noted
that these emission impacts are not due to the new GHG vehicle
standards. These impacts are instead a consequence of the assumed
ethanol volumes. This program does not mandate an increase in E10, nor
any particular fuel blend. The emission impact of this shift was also
modeled in the RFS2 rule.
As shown in Table III.G-1, EPA estimates that this program would
result in reductions of NOX, VOC, PM and

[[Page 25497]]

SOX, but would increase CO emissions. For NOX,
VOC, and PM we estimate net reductions because the emissions reductions
from upstream sources are larger than the emission increases due to
downstream sources. In the case of CO, we estimate slight emission
increases, because there are relatively small reductions in upstream
emissions, and thus the projected downstream emission increases are
greater than the projected emission decreases due to reduced fuel
production. For SOX, downstream emissions are roughly
proportional to fuel consumption, therefore a decrease is seen in both
upstream and downstream sources.
For all criteria pollutants the overall impact of the program would
be relatively small compared to total U.S. inventories across all
sectors. In 2030, EPA estimates the program would reduce total
NOX, PM and SOX inventories by 0.1 to 0.8 percent
and reduce the VOC inventory by 1.0 percent, while increasing the total
national CO inventory by 0.6 percent.
As shown in Table III.G-2, EPA estimates that the GHG program would
result in small changes for air toxic emissions compared to total U.S.
inventories across all sectors. In 2030, EPA estimates the program
would reduce total benzene and 1,3 butadiene emissions by 0.1 to 0.3
percent. Total acrolein and formaldehyde emissions would increase by
0.1 percent. Acetaldehyde emissions would increase by 2.2 percent.
One commenter requested that EPA present emission inventories for
additional air toxics. EPA is presenting inventories for certain air
toxic emissions which were identified as key national and regional-
scale cancer and noncancer risk drivers in past National Air Toxics
Assessments (NATA). For additional details, please refer to the
Response to Comments document.\327\
---------------------------------------------------------------------------

\327\ U.S. EPA. National Air Toxics Assessment. 2002, 1999, and
1996. Available at: http://www.epa.gov/nata/.
---------------------------------------------------------------------------

Other factors which may impact non-GHG emissions, but are not
estimated in this analysis, include:
Vehicle technologies used to reduce tailpipe
CO2 emissions; because the regulatory standards for non-GHG
emissions are the primary driver for these emissions, EPA expects the
impact of this program to be negligible on non-GHG emission rates per
mile.
The potential for increased market penetration of diesel
vehicles; because these vehicles would be held to the same
certification and in-use standards for criteria pollutants as their
gasoline counterparts, EPA expects their impact to be negligible on
criteria pollutants and other non-GHG emissions. EPA does not project
increased penetration of diesels as necessary to meet the GHG
standards.
Early introduction of electric vehicles and plug-in hybrid
electric vehicles, which would reduce criteria emissions in cases where
those vehicles are able to be certified to lower certification
standards. This would also likely reduce gaseous air toxics.
Reduced refueling emissions due to less frequent refueling
events and reduced annual refueling volumes resulting from the GHG
standards.
Increased hot soak evaporative emissions due to the likely
increase in number of trips associated with VMT rebound modeled in this
rule.

Table III.G-1--Annual Criteria Emission Impacts of Program
[Short tons]
----------------------------------------------------------------------------------------------------------------
Total impacts Upstream impacts Downstream impacts
-----------------------------------------------------------------------------
2020 2030 2020 2030 2020 2030
----------------------------------------------------------------------------------------------------------------
VOC............................... -60,187 -115,542 -64,506 -126,749 4,318 11,207
% of total inventory.......... -0.51% -1.01% -0.55% -1.11% 0.04% 0.01%
CO................................ 3,992 170,675 -6,165 -12,113 10,156 182,788
% of total inventory.......... 0.01% 0.56% -0.02% -0.04% 0.01% 0.6%
NOX............................... -5,881 -21,763 -19,291 -37,905 13,410 16,143
% of total inventory.......... -0.02 -0.07% -0.06% -0.12% 0.04% 0.05%
PM2.5............................. -2,398 -4,564 -2,629 -5,165 231.0 602.3
% of total inventory.......... -0.03% -0.05% -0.03% -0.06% 0.00% 0.01%
SOX............................... -13,832 -27,443 -11,804 -23,194 -2,027 -4,249
% of total inventory.......... -0.41% -0.82% -0.35% -0.69% -0.06% -0.13%
----------------------------------------------------------------------------------------------------------------

Table III.G-2--Annual Air Toxic Emission Impacts of Program
[Short tons]
----------------------------------------------------------------------------------------------------------------
Total impacts Upstream impacts Downstream impacts
-----------------------------------------------------------------------------
2020 2030 2020 2030 2020 2030
----------------------------------------------------------------------------------------------------------------
1,3[dash]Butadiene................ -95 -21 -1.5 -3.0 -93.6 -18.1
% of total inventory.......... -0.38% -0.10% -0.01% -0.01% -0.37% -0.09%
Acetaldehyde...................... 760 668 -6.8 -13.4 766.9 681.5
% of total inventory.......... 2.26% 2.18% -0.02% -0.04% 2.28% 2.22%
Acrolein.......................... 1 5 -0.9 -1.8 1.7 6.5
% of total inventory.......... 0.01% 0.07% -0.01% -0.03% 0.03% 0.10%
Benzene........................... -890 -523 -139.6 -274.3 -750.0 -248.3
% of total inventory.......... -0.48% -0.29% -0.08% -0.15% -0.40% -0.14%
Formaldehyde...................... -49 15 -51.4 -101.0 2.1 116.3
% of total inventory.......... -0.06% 0.02% -0.06% -0.12% 0.00% 0.14%
----------------------------------------------------------------------------------------------------------------

[[Page 25498]]

1. Upstream Impacts of Program
No substantive comments were received on the upstream inventory
modeling used in the proposal. The rulemaking inventories were updated
with the revised estimates of fuel savings as detailed in Section
III.F.
Reducing tailpipe CO2 emissions from light-duty cars and
trucks through tailpipe standards and improved A/C efficiency will
result in reduced fuel demand and reductions in the emissions
associated with all of the processes involved in getting petroleum to
the pump. These upstream emission impacts on criteria pollutants are
summarized in Table III.G-1. The upstream reductions grow over time as
the fleet turns over to cleaner CO2 vehicles, so that by
2030 VOC would decrease by 127,000 tons, NOX by 38,000 tons,
and PM2.5 by 5,000 tons. Table III.G-2 shows the
corresponding impacts on upstream air toxic emissions in 2030.
Formaldehyde decreases by 101 tons, benzene by 274 tons, acetaldehyde
by 13 tons, acrolein by 2 tons, and 1,3-butadiene by 3 tons.
To determine these impacts, EPA estimated the impact of reduced
petroleum volumes on the extraction and transportation of crude oil as
well as the production and distribution of finished gasoline. For the
purpose of assessing domestic-only emission reductions it was necessary
to estimate the fraction of fuel savings attributable to domestic
finished gasoline, and of this gasoline what fraction is produced from
domestic crude. For this analysis EPA estimated that 50 percent of fuel
savings is attributable to domestic finished gasoline and that 90
percent of this gasoline originated from imported crude. Emission
factors for most upstream emission sources are based on the GREET1.8
model, developed by DOE's Argonne National Laboratory,\328\ but in some
cases the GREET values were modified or updated by EPA to be consistent
with the National Emission Inventory (NEI).\329\ The primary updates
for this analysis were to incorporate newer information on gasoline
distribution emissions for VOC from the NEI, which were significantly
higher than GREET estimates; and the incorporation of upstream emission
factors for the air toxics estimated in this analysis: benzene, 1,3-
butadiene, acetaldehyde, acrolein, and formaldehyde. The development of
these emission factors is detailed in RIA Chapter 5.
---------------------------------------------------------------------------

\328\ Greenhouse Gas, Regulated Emissions, and Energy Use in
Transportation model (GREET), U.S. Department of Energy, Argonne
National Laboratory, http://www.transportation.anl.gov/modeling_
simulation/GREET/.
\329\ U.S. EPA. 2002 National Emissions Inventory (NEI) Data and
Documentation, http://www.epa.gov/ttn/chief/net/2002inventory.html.
---------------------------------------------------------------------------

2. Downstream Impacts of Program
No substantive comments were received on the emission modeling or
emission inventories presented in this section. However, two changes in
modeling differentiate the analysis presented here from that presented
in the proposal. Economic inputs such as fuel prices and vehicle sales
were updated from AEO 2009 to AEO 2010 Early Release, and as described
above, the effects of ethanol volume assumptions were explicitly
modeled. Thus, the primary differences in non-GHG emissions between the
proposed rule and final rule are attributed more to these changes in
analytic inputs, and less to changes in the GHG standards program.
Downstream emission impacts attributable to this program are due to
the VMT rebound effect and the ethanol volume assumptions. As discussed
in more detail in Section III.H, the effect of fuel cost on VMT
(``rebound'') was accounted for in our assessment of economic and
environmental impacts of this rule. A 10 percent rebound case was used
for this analysis, meaning that VMT for affected model years is modeled
as increasing by 10 percent as much as the increase in fuel economy;
i.e., a 10 percent increase in fuel economy would yield approximately a
1 percent increase in VMT.
As detailed in the introduction to this section, fuel composition
also has effects on vehicle emissions and particularly air toxics. The
relationship between fuel composition and emission impacts used in
MOVES2010 and applied in this analysis match those developed for the
recent Renewable Fuels Standard (RFS2) requirement, and are extensively
documented in the RFS2 RIA and supporting documents.\330\
---------------------------------------------------------------------------

\330\ EPA 2010, Renewable Fuel Standard Program (RFS2)
Regulatory Impact Analysis. EPA-420-R-10-006. February 2010. Docket
EPA-HQ-OAR-2009-0472-11332. See also 75 FR 14670, March 26, 2010.
---------------------------------------------------------------------------

Downstream emission impacts of the rebound effect are summarized in
Table III.G-1 for criteria pollutants and precursors and Table III.G-2
for air toxics. The emission impacts from the rebound effect and the
change in fuel supply grow over time as the fleet turns over to cleaner
CO2 vehicles, so that by 2030 VOC would increase by 11,000
tons, NOX by 16,000 tons, and PM2.5 by 600 tons.
Table III.G-2 shows the corresponding impacts on air toxic emissions.
These impacts in 2030 include 18 fewer tons of 1,3-butadiene, 668
additional tons of acetaldehyde, 248 fewer tons of benzene, 116
additional tons of formaldehyde, and 6.5 additional tons of acrolein.
For this analysis, MOVES2010 was used to estimate base VOC, CO,
NOX, PM and air toxics emissions for both control and
reference cases. Rebound emissions from light duty cars and trucks were
then calculated using the OMEGA model post-processor and added to the
control case. A more complete discussion of the inputs, methodology,
and results is contained in RIA Chapter 5.
3. Health Effects of Non-GHG Pollutants
In this section we discuss health effects associated with exposure
to some of the criteria and air toxics impacted by the vehicle
standards; PM, ozone, NOX and SOX, CO and air
toxics. No substantive comments were received on the health effects of
non-GHG pollutants.
a. Particulate Matter
i. Background
Particulate matter is a generic term for a broad class of
chemically and physically diverse substances. It can be principally
characterized as discrete particles that exist in the condensed (liquid
or solid) phase spanning several orders of magnitude in size. Since
1987, EPA has delineated that subset of inhalable particles small
enough to penetrate to the thoracic region (including the
tracheobronchial and alveolar regions) of the respiratory tract
(referred to as thoracic particles). Current NAAQS use PM2.5
as the indicator for fine particles (with PM2.5 referring to
particles with a nominal mean aerodynamic diameter less than or equal
to 2.5 [mu]m), and use PM10 as the indicator for purposes of
regulating the coarse fraction of PM10 (referred to as
thoracic coarse particles or coarse-fraction particles; generally
including particles with a nominal mean aerodynamic diameter greater
than 2.5 [mu]m and less than or equal to 10 [mu]m, or
PM10-2.5). Ultrafine particles are a subset of fine
particles, generally less than 100 nanometers (0.1 [mu]m) in
aerodynamic diameter.
Fine particles are produced primarily by combustion processes and
by transformations of gaseous emissions (e.g., SOX,
NOX and VOC) in the atmosphere. The chemical and physical
properties of PM2.5 may vary greatly with time, region,
meteorology, and source category. Thus, PM2.5 may include a
complex mixture of different pollutants including sulfates, nitrates,
organic compounds, elemental carbon

[[Page 25499]]

and metal compounds. These particles can remain in the atmosphere for
days to weeks and travel hundreds to thousands of kilometers.
ii. Health Effects of PM
Scientific studies show ambient PM is associated with a series of
adverse health effects. These health effects are discussed in detail in
EPA's Integrated Science Assessment for Particulate Matter (ISA).\331\
Further discussion of health effects associated with PM can also be
found in the RIA for this rule. The ISA summarizes evidence associated
with PM2.5, PM10-2.5, and ultrafine particles
(UFPs).
---------------------------------------------------------------------------

\331\ U.S. EPA (2009) Integrated Science Assessment for
Particulate Matter. EPA 600/R-08/139F, Docket EPA-HQ-OAR-2009-0472-
11295.
---------------------------------------------------------------------------

The ISA concludes that health effects associated with short-term
exposures (hours to days) to ambient PM2.5 include non-fatal
cardiovascular effects, mortality, and respiratory effects, such as
exacerbation of asthma symptoms in children and hospital admissions and
emergency department visits for chronic obstructive pulmonary disease
(COPD) and respiratory infections.\332\ The ISA notes that long-term
exposure to PM2.5 (months to years) is associated with the
development/progression of cardiovascular disease, premature mortality,
and respiratory effects, including reduced lung function growth,
increased respiratory symptoms, and asthma development.\333\ The ISA
concludes that that the currently available scientific evidence from
epidemiologic, controlled human exposure studies, and toxicological
studies supports that a causal association exists between short- and
long-term exposures to PM2.5 and cardiovascular effects and
mortality. Furthermore, the ISA concludes that the collective evidence
supports likely causal associations between short- and long-term
PM2.5 exposures and respiratory effects. The ISA also
concludes that the evidence is suggestive of a causal association for
reproductive and developmental effects and cancer, mutagenicity, and
genotoxicity and long-term exposure to PM2.5.\334\
---------------------------------------------------------------------------

\332\ U.S. EPA (2009). Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.1.1.
\333\ U.S. EPA (2009). Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-08/139F, 2009. page 2-12, Sections
7.3.1.1 and 7.3.2.1.
\334\ U.S. EPA (2009). Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.2.
---------------------------------------------------------------------------

For PM10-2.5, the ISA concludes that the current
evidence is suggestive of a causal relationship between short-term
exposures and cardiovascular effects, such as hospitalization for
ischemic heart disease. There is also suggestive evidence of a causal
relationship between short-term PM10-2.5 exposure and
mortality and respiratory effects. Data are inadequate to draw
conclusions regarding the health effects associated with long-term
exposure to PM10-2.5.\335\
---------------------------------------------------------------------------

\335\ U.S. EPA (2009). Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.4,
Table 2-6.
---------------------------------------------------------------------------

For UFPs, the ISA concludes that there is suggestive evidence of a
causal relationship between short-term exposures and cardiovascular
effects, such as changes in heart rhythm and blood vessel function. It
also concludes that there is suggestive evidence of association between
short-term exposure to UFPs and respiratory effects. Data are
inadequate to draw conclusions regarding the health effects associated
with long-term exposure to UFP's.\336\
---------------------------------------------------------------------------

\336\ U.S. EPA (2009). Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.5,
Table 2-6.
---------------------------------------------------------------------------

b. Ozone
i. Background
Ground-level ozone pollution is typically formed by the reaction of
VOC and NOX in the lower atmosphere in the presence of heat
and sunlight. These pollutants, often referred to as ozone precursors,
are emitted by many types of pollution sources, such as highway and
nonroad motor vehicles and engines, power plants, chemical plants,
refineries, makers of consumer and commercial products, industrial
facilities, and smaller area sources.
The science of ozone formation, transport, and accumulation is
complex.\337\ Ground-level ozone is produced and destroyed in a
cyclical set of chemical reactions, many of which are sensitive to
temperature and sunlight. When ambient temperatures and sunlight levels
remain high for several days and the air is relatively stagnant, ozone
and its precursors can build up and result in more ozone than typically
occurs on a single high-temperature day. Ozone can be transported
hundreds of miles downwind from precursor emissions, resulting in
elevated ozone levels even in areas with low local VOC or
NOX emissions.
---------------------------------------------------------------------------

\337\ U.S. EPA (2006). Air Quality Criteria for Ozone and
Related Photochemical Oxidants (Final). EPA/600/R-05/004aF-cF.
Washington, DC: U.S. EPA. Docket EPA-HQ-OAR-2009-0472-0099 through -
0101.
---------------------------------------------------------------------------

ii. Health Effects of Ozone
The health and welfare effects of ozone are well documented and are
assessed in EPA's 2006 Air Quality Criteria Document (ozone AQCD) and
2007 Staff Paper.^{338 339} Ozone can irritate the respiratory
system, causing coughing, throat irritation, and/or uncomfortable
sensation in the chest. Ozone can reduce lung function and make it more
difficult to breathe deeply; breathing may also become more rapid and
shallow than normal, thereby limiting a person's activity. Ozone can
also aggravate asthma, leading to more asthma attacks that require
medical attention and/or the use of additional medication. In addition,
there is suggestive evidence of a contribution of ozone to
cardiovascular-related morbidity and highly suggestive evidence that
short-term ozone exposure directly or indirectly contributes to non-
accidental and cardiopulmonary-related mortality, but additional
research is needed to clarify the underlying mechanisms causing these
effects. In a recent report on the estimation of ozone-related
premature mortality published by the National Research Council (NRC), a
panel of experts and reviewers concluded that short-term exposure to
ambient ozone is likely to contribute to premature deaths and that
ozone-related mortality should be included in estimates of the health
benefits of reducing ozone exposure.\340\ Animal toxicological evidence
indicates that with repeated exposure, ozone can inflame and damage the
lining of the lungs, which may lead to permanent changes in lung tissue
and irreversible reductions in lung function. People who are more
susceptible to effects associated with exposure to ozone can include
children, the elderly, and individuals with respiratory disease such as
asthma. Those with greater exposures to ozone, for instance due to

[[Page 25500]]

time spent outdoors (e.g., children and outdoor workers), are of
particular concern.
---------------------------------------------------------------------------

\338\ U.S. EPA. (2006). Air Quality Criteria for Ozone and
Related Photochemical Oxidants (Final). EPA/600/R-05/004aF-cF.
Washington, DC: U.S. EPA.
\339\ U.S. EPA (2007). Review of the National Ambient Air
Quality Standards for Ozone: Policy Assessment of Scientific and
Technical Information, OAQPS Staff Paper. EPA-452/R-07-003.
Washington, DC, U.S. EPA. Docket EPA-HQ-OAR-2009-0472-0105 through -
0106.
\340\ National Research Council (NRC), 2008. Estimating
Mortality Risk Reduction and Economic Benefits from Controlling
Ozone Air Pollution. The National Academies Press: Washington, DC
Docket EPA-HQ-OAR-2009-0472-0322.
---------------------------------------------------------------------------

The 2006 ozone AQCD also examined relevant new scientific
information that has emerged in the past decade, including the impact
of ozone exposure on such health effects as changes in lung structure
and biochemistry, inflammation of the lungs, exacerbation and causation
of asthma, respiratory illness-related school absence, hospital
admissions and premature mortality. Animal toxicological studies have
suggested potential interactions between ozone and PM with increased
responses observed to mixtures of the two pollutants compared to either
ozone or PM alone. The respiratory morbidity observed in animal studies
along with the evidence from epidemiologic studies supports a causal
relationship between acute ambient ozone exposures and increased
respiratory-related emergency room visits and hospitalizations in the
warm season. In addition, there is suggestive evidence of a
contribution of ozone to cardiovascular-related morbidity and non-
accidental and cardiopulmonary mortality.
c. NOX and SOX
i. Background
Nitrogen dioxide (NO2) is a member of the NOX
family of gases. Most NO2 is formed in the air through the
oxidation of nitric oxide (NO) emitted when fuel is burned at a high
temperature. SO2, a member of the sulfur oxide
(SOX) family of gases, is formed from burning fuels
containing sulfur (e.g., coal or oil derived), extracting gasoline from
oil, or extracting metals from ore.
SO2 and NO2 can dissolve in water vapor and
further oxidize to form sulfuric and nitric acid which react with
ammonia to form sulfates and nitrates, both of which are important
components of ambient PM. The health effects of ambient PM are
discussed in Section III.G.3.a of this preamble. NOX along
with non-methane hydrocarbon (NMHC) are the two major precursors of
ozone. The health effects of ozone are covered in Section III.G.3.b.
ii. Health Effects of NO2
Information on the health effects of NO2 can be found in
the EPA Integrated Science Assessment (ISA) for Nitrogen Oxides.\341\
The EPA has concluded that the findings of epidemiologic, controlled
human exposure, and animal toxicological studies provide evidence that
is sufficient to infer a likely causal relationship between respiratory
effects and short-term NO2 exposure. The ISA concludes that
the strongest evidence for such a relationship comes from epidemiologic
studies of respiratory effects including symptoms, emergency department
visits, and hospital admissions. The ISA also draws two broad
conclusions regarding airway responsiveness following NO2
exposure. First, the ISA concludes that NO2 exposure may
enhance the sensitivity to allergen-induced decrements in lung function
and increase the allergen-induced airway inflammatory response
following 30-minute exposures of asthmatics to NO2
concentrations as low as 0.26 ppm. In addition, small but significant
increases in non-specific airway hyperresponsiveness were reported
following 1-hour exposures of asthmatics to 0.1 ppm NO2.
Second, exposure to NO2 has been found to enhance the
inherent responsiveness of the airway to subsequent nonspecific
challenges in controlled human exposure studies of asthmatic subjects.
Enhanced airway responsiveness could have important clinical
implications for asthmatics since transient increases in airway
responsiveness following NO2 exposure have the potential to
increase symptoms and worsen asthma control. Together, the
epidemiologic and experimental data sets form a plausible, consistent,
and coherent description of a relationship between NO2
exposures and an array of adverse health effects that range from the
onset of respiratory symptoms to hospital admission.
---------------------------------------------------------------------------

\341\ U.S. EPA (2008). Integrated Science Assessment for Oxides
of Nitrogen--Health Criteria (Final Report). EPA/600/R-08/071.
Washington, DC: U.S.EPA. Docket EPA-HQ-OAR-2009-0472-0350.
---------------------------------------------------------------------------

Although the weight of evidence supporting a causal relationship is
somewhat less certain than that associated with respiratory morbidity,
NO2 has also been linked to other health endpoints. These
include all-cause (nonaccidental) mortality, hospital admissions or
emergency department visits for cardiovascular disease, and decrements
in lung function growth associated with chronic exposure.
iii. Health Effects of SO2
Information on the health effects of SO2 can be found in
the EPA Integrated Science Assessment for Sulfur Oxides.\342\
SO2 has long been known to cause adverse respiratory health
effects, particularly among individuals with asthma. Other potentially
sensitive groups include children and the elderly. During periods of
elevated ventilation, asthmatics may experience symptomatic
bronchoconstriction within minutes of exposure. Following an extensive
evaluation of health evidence from epidemiologic and laboratory
studies, the EPA has concluded that there is a causal relationship
between respiratory health effects and short-term exposure to
SO2. Separately, based on an evaluation of the epidemiologic
evidence of associations between short-term exposure to SO2
and mortality, the EPA has concluded that the overall evidence is
suggestive of a causal relationship between short-term exposure to
SO2 and mortality.
---------------------------------------------------------------------------

\342\ U.S. EPA. (2008). Integrated Science Assessment (ISA) for
Sulfur Oxides--Health Criteria (Final Report). EPA/600/R-08/047F.
Washington, DC: U.S. Environmental Protection Agency. Docket EPA-HQ-
OAR-2009-0472-0335.
---------------------------------------------------------------------------

d. Carbon Monoxide
Information on the health effects of carbon monoxide (CO) can be
found in the EPA Integrated Science Assessment (ISA) for Carbon
Monoxide.\343\ The ISA concludes that ambient concentrations of CO are
associated with a number of adverse health effects.\344\ This section
provides a summary of the health effects associated with exposure to
ambient concentrations of CO.\345\
---------------------------------------------------------------------------

\343\ U.S. EPA, 2010. Integrated Science Assessment for Carbon
Monoxide (Final Report). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-09/019F, 2010. http://cfpub.epa.gov/ncea/
cfm/recordisplay.cfm?deid=218686.
\344\ The ISA evaluates the health evidence associated with
different health effects, assigning one of five ``weight of
evidence'' determination: causal relationship, likely to be a causal
relationship, suggestive of a causal relationship, inadequate to
infer a causal relationship, and not likely to be a causal
relationship. For definitions of these levels of evidence, please
refer to Section 1.6 of the ISA.
\345\ Personal exposure includes contributions from many
sources, and in many different environments. Total personal exposure
to CO includes both ambient and nonambient components; and both
components may contribute to adverse health effects.
---------------------------------------------------------------------------

Human clinical studies of subjects with coronary artery disease
show a decrease in the time to onset of exercise-induced angina (chest
pain) and electrocardiogram changes following CO exposure. In addition,
epidemiologic studies show associations between short-term CO exposure
and cardiovascular morbidity, particularly increased emergency room
visits and hospital admissions for coronary heart disease (including
ischemic heart disease, myocardial infarction, and angina). Some
epidemiologic evidence is also available for increased hospital
admissions and emergency room visits for congestive heart failure and
cardiovascular disease as a whole. The ISA concludes that a causal
relationship is likely between short-term exposures to CO and
cardiovascular morbidity. It also concludes that available data are
inadequate to conclude that a causal

[[Page 25501]]

relationship exists between long-term exposures to CO and
cardiovascular morbidity.
Animal studies show various neurological effects with in-utero CO
exposure. Controlled human exposure studies report inconsistent neural
and behavioral effects following low-level CO exposures. The ISA
concludes the evidence is suggestive of a causal relationship with both
short- and long-term exposure to CO and central nervous system effects.
A number of epidemiologic and animal toxicological studies cited in
the ISA have evaluated associations between preterm birth and cardiac
birth defects and CO exposure. The epidemiologic studies provide
limited evidence of a CO-induced effect on pre-term births and birth
defects, with weak evidence for a decrease in birth weight. Animal
toxicological studies have found associations between perinatal CO
exposure and decrements in birth weight, as well as other developmental
outcomes. The ISA concludes these studies are suggestive of a causal
relationship between long-term exposures to CO and developmental
effects and birth outcomes.
Epidemiologic studies provide evidence of effects on respiratory
morbidity such as changes in pulmonary function, respiratory symptoms,
and hospital admissions associated with ambient CO concentrations. A
limited number of epidemiologic studies considered copollutants such as
ozone, SO2, and PM in two-pollutant models and found that CO
risk estimates were generally robust, although this limited evidence
makes it difficult to disentangle effects attributed to CO itself from
those of the larger complex air pollution mixture. Controlled human
exposure studies have not extensively evaluated the effect of CO on
respiratory morbidity. Animal studies at levels of 50-100 ppm CO show
preliminary evidence of altered pulmonary vascular remodeling and
oxidative injury. The ISA concludes that the evidence is suggestive of
a causal relationship between short-term CO exposure and respiratory
morbidity, and inadequate to conclude that a causal relationship exists
between long-term exposure and respiratory morbidity.
Finally, the ISA concludes that the epidemiologic evidence is
suggestive of a causal relationship between short-term exposures to CO
and mortality. Epidemiologic studies provide evidence of an association
between short-term exposure to CO and mortality, but limited evidence
is available to evaluate cause-specific mortality outcomes associated
with CO exposure. In addition, the attenuation of CO risk estimates
which was often observed in copollutant models contributes to the
uncertainty as to whether CO is acting alone or as an indicator for
other combustion-related pollutants. The ISA also concludes that there
is not likely to be a causal relationship between relevant long-term
exposures to CO and mortality.
e. Air Toxics
Motor vehicle emissions contribute to ambient levels of air toxics
known or suspected as human or animal carcinogens, or that have
noncancer health effects. The population experiences an elevated risk
of cancer and other noncancer health effects from exposure to the class
of pollutants known collectively as ``air toxics''.\346\ These
compounds include, but are not limited to, benzene, 1,3-butadiene,
formaldehyde, acetaldehyde, acrolein, polycyclic organic matter (POM),
and naphthalene. These compounds, except acetaldehyde, were identified
as national or regional risk drivers in the 2002 National-scale Air
Toxics Assessment (NATA) and have significant inventory contributions
from mobile sources.\347\ Emissions and ambient concentrations of
compounds are discussed in the RIA chapters on emission inventories and
air quality (Chapters 5 and 7, respectively).
---------------------------------------------------------------------------

\346\ U.S. EPA. 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata12002/risksum.html. Docket EPA-HQ-
OAR-2009-0472-11322.
\347\ U.S. EPA. 2009. National-Scale Air Toxics Assessment for
2002. http://www.epa.gov/ttn/atw/nata2002/. Docket EPA-HQ-OAR-2009-
0472-11321.
---------------------------------------------------------------------------

i. Benzene
The EPA's IRIS database lists benzene as a known human carcinogen
(causing leukemia) by all routes of exposure, and concludes that
exposure is associated with additional health effects, including
genetic changes in both humans and animals and increased proliferation
of bone marrow cells in mice.^{348 349 350} EPA states in its
IRIS database that data indicate a causal relationship between benzene
exposure and acute lymphocytic leukemia and suggest a relationship
between benzene exposure and chronic non-lymphocytic leukemia and
chronic lymphocytic leukemia. The International Agency for Research on
Carcinogens (IARC) has determined that benzene is a human carcinogen
and the U.S. Department of Health and Human Services (DHHS) has
characterized benzene as a known human carcinogen.^{351 352}
---------------------------------------------------------------------------

\348\ U.S. EPA. 2000. Integrated Risk Information System File
for Benzene. This material is available electronically at http://
www.epa.gov/iris/subst/0276.htm. Docket EPA-HQ-OAR-2009-0472-1659.
\349\ International Agency for Research on Cancer (IARC). 1982.
Monographs on the evaluation of carcinogenic risk of chemicals to
humans, Volume 29. Some industrial chemicals and dyestuffs, World
Health Organization, Lyon, France, p. 345-389. Docket EPA-HQ-OAR-
2009-0472-0366.
\350\ Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry,
V.A. 1992. Synergistic action of the benzene metabolite hydroquinone
on myelopoietic stimulating activity of granulocyte/macrophage
colony-stimulating factor in vitro, Proc. Natl. Acad. Sci. 89:3691-
3695. Docket EPA-HQ-OAR-2009-0472-0370.
\351\ International Agency for Research on Cancer (IARC). 1982.
Monographs on the evaluation of carcinogenic risk of chemicals to
humans, Volume 29. Some industrial chemicals and dyestuffs, World
Health Organization, Lyon, France. Docket EPA-HQ-OAR-2009-0472-0366.
\352\ U.S. Department of Health and Human Services National
Toxicology Program 11th Report on Carcinogens available at: http://
ntp.niehs.nih.gov/go/16183.
---------------------------------------------------------------------------

A number of adverse noncancer health effects including blood
disorders, such as preleukemia and aplastic anemia, have also been
associated with long-term exposure to benzene.^{353 354} The
most sensitive noncancer effect observed in humans, based on current
data, is the depression of the absolute lymphocyte count in
blood.^{355 356} In addition, recent work, including studies
sponsored by the Health Effects Institute (HEI), provides evidence that
biochemical responses are occurring at lower levels of benzene exposure
than previously known.^{357 358 359 360} EPA's

[[Page 25502]]

IRIS program has not yet evaluated these new data.
---------------------------------------------------------------------------

\353\ Aksoy, M. (1989). Hematotoxicity and carcinogenicity of
benzene. Environ. Health Perspect. 82: 193-197. Docket EPA-HQ-OAR-
2009-0472-0368.
\354\ Goldstein, B.D. (1988). Benzene toxicity. Occupational
medicine. State of the Art Reviews. 3: 541-554. Docket EPA-HQ-OAR-
2009-0472-0325.
\355\ Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E.
Marti, Y.Z. Wang, M. Linet, L.Q. Xi, W. Lu, M.T. Smith, N. Titenko-
Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B. Hayes (1996)
Hematotoxicity among Chinese workers heavily exposed to benzene. Am.
J. Ind. Med. 29: 236-246. Docket EPA-HQ-OAR-2009-0472-0326.
\356\ U.S. EPA (2002) Toxicological Review of Benzene (Noncancer
Effects). Environmental Protection Agency, Integrated Risk
Information System (IRIS), Research and Development, National Center
for Environmental Assessment, Washington DC. This material is
available electronically at http://www.epa.gov/iris/subst/0276.htm.
Docket EPA-HQ-OAR-2009-0472-0327.
\357\ Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, B.;
Melikian, A.; Eastmond, D.; Rappaport, S.; Li, H.; Rupa, D.;
Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.; Winnik, M.; Kwok,
E.; Li, Y.; Mu, R.; Xu, B.; Zhang, X.; Li, K. (2003) HEI Report 115,
Validation & Evaluation of Biomarkers in Workers Exposed to Benzene
in China. Docket EPA-HQ-OAR-2009-0472-0328.
\358\ Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et
al. (2002) Hematological changes among Chinese workers with a broad
range of benzene exposures. Am. J. Industr. Med. 42: 275-285. Docket
EPA-HQ-OAR-2009-0472-0329.
\359\ Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004)
Hematotoxically in Workers Exposed to Low Levels of Benzene. Science
306: 1774-1776. Docket EPA-HQ-OAR-2009-0472-0330.
\360\ Turtletaub, K.W. and Mani, C. (2003) Benzene metabolism in
rodents at doses relevant to human exposure from Urban Air. Research
Reports Health Effect Inst. Report No.113. Docket EPA-HQ-OAR-2009-
0472-0385.
---------------------------------------------------------------------------

ii. 1,3-Butadiene
EPA has characterized 1,3-butadiene as carcinogenic to humans by
inhalation.^{361 362} The IARC has determined that 1,3-
butadiene is a human carcinogen and the U.S. DHHS has characterized
1,3-butadiene as a known human carcinogen.^{363 364} There are
numerous studies consistently demonstrating that 1,3-butadiene is
metabolized into genotoxic metabolites by experimental animals and
humans. The specific mechanisms of 1,3-butadiene-induced carcinogenesis
are unknown; however, the scientific evidence strongly suggests that
the carcinogenic effects are mediated by genotoxic metabolites. Animal
data suggest that females may be more sensitive than males for cancer
effects associated with 1,3-butadiene exposure; there are insufficient
data in humans from which to draw conclusions about sensitive
subpopulations. 1,3-butadiene also causes a variety of reproductive and
developmental effects in mice; no human data on these effects are
available. The most sensitive effect was ovarian atrophy observed in a
lifetime bioassay of female mice.\365\
---------------------------------------------------------------------------

\361\ U.S. EPA (2002) Health Assessment of 1,3-Butadiene. Office
of Research and Development, National Center for Environmental
Assessment, Washington Office, Washington, DC. Report No. EPA600-P-
98-001F. This document is available electronically at http://
www.epa.gov/iris/supdocs/buta-sup.pdf. Docket EPA-HQ-OAR-2009-0472-
0386.
\362\ U.S. EPA (2002) Full IRIS Summary for 1,3-butadiene (CASRN
106-99-0). Environmental Protection Agency, Integrated Risk
Information System (IRIS), Research and Development, National Center
for Environmental Assessment, Washington, DC. http://www.epa.gov/
iris/subst/0139.htm. Docket EPA-HQ-OAR-2009-0472-1660
\363\ International Agency for Research on Cancer (IARC) (1999)
Monographs on the evaluation of carcinogenic risk of chemicals to
humans, Volume 71, Re-evaluation of some organic chemicals,
hydrazine and hydrogen peroxide and Volume 97 (in preparation),
World Health Organization, Lyon, France. Docket EPA-HQ-OAR-2009-
0472-0387.
\364\ U.S. Department of Health and Human Services (2005)
National Toxicology Program 11th Report on Carcinogens available at:
ntp.niehs.nih.gov/index.cfm?objectid=32BA9724-F1F6-975E-
7FCE50709CB4C932.
\365\ Bevan, C.; Stadler, J.C.; Elliot, G.S.; et al. (1996)
Subchronic toxicity of 4-vinylcyclohexene in rats and mice by
inhalation. Fundam. Appl. Toxicol. 32:1-10. Docket EPA-HQ-OAR-2009-
0472-0388.
---------------------------------------------------------------------------

iii. Formaldehyde
Since 1987, EPA has classified formaldehyde as a probable human
carcinogen based on evidence in humans and in rats, mice, hamsters, and
monkeys.\366\ EPA is currently reviewing recently published
epidemiological data. For instance, research conducted by the National
Cancer Institute (NCI) found an increased risk of nasopharyngeal cancer
and lymphohematopoietic malignancies such as leukemia among workers
exposed to formaldehyde.^{367 368} In an analysis of the
lymphohematopoietic cancer mortality from an extended follow-up of
these workers, NCI confirmed an association between lymphohematopoietic
cancer risk and peak exposures.\369\ A recent National Institute of
Occupational Safety and Health (NIOSH) study of garment workers also
found increased risk of death due to leukemia among workers exposed to
formaldehyde.\370\ Extended follow-up of a cohort of British chemical
workers did not find evidence of an increase in nasopharyngeal or
lymphohematopoietic cancers, but a continuing statistically significant
excess in lung cancers was reported.\371\ Recently, the IARC re-
classified formaldehyde as a human carcinogen (Group 1).\372\
---------------------------------------------------------------------------

\366\ U.S. EPA (1987) Assessment of Health Risks to Garment
Workers and Certain Home Residents from Exposure to Formaldehyde,
Office of Pesticides and Toxic Substances, April 1987. Docket EPA-
HQ-OAR-2009-0472-0389.
\367\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.;
Blair, A. 2003. Mortality from lymphohematopoetic malignancies among
workers in formaldehyde industries. Journal of the National Cancer
Institute 95: 1615-1623. Docket EPA-HQ-OAR-2009-0472-0336.
\368\ Hauptmann, M..; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.;
Blair, A. 2004. Mortality from solid cancers among workers in
formaldehyde industries. American Journal of Epidemiology 159: 1117-
1130. Docket EPA-HQ-OAR-2009-0472-0337.
\369\ Beane Freeman, L.E.; Blair, A.; Lubin, J.H.; Stewart,
P.A.; Hayes, R.B.; Hoover, R.N.; Hauptmann, M. 2009. Mortality from
lymphohematopoietic malignancies among workers in formaldehyde
industries: The National Cancer Institute cohort. J. National Cancer
Inst. 101: 751-761. Docket EPA-HQ-OAR-2009-0472-0338.
\370\ Pinkerton, L.E. 2004. Mortality among a cohort of garment
workers exposed to formaldehyde: an update. Occup. Environ. Med. 61:
193-200. Docket EPA-HQ-OAR-2009-0472-0339.
\371\ Coggon, D, EC Harris, J Poole, KT Palmer. 2003. Extended
follow-up of a cohort of British chemical workers exposed to
formaldehyde. J National Cancer Inst. 95:1608-1615. Docket EPA-HQ-
OAR-2009-0472-0340.
\372\ International Agency for Research on Cancer (IARC). 2006.
Formaldehyde, 2-Butoxyethanol and 1-tert-Butoxypropan-2-ol. Volume
88. (in preparation), World Health Organization, Lyon, France.
Docket EPA-HQ-OAR-2009-0472-1164.
---------------------------------------------------------------------------

Formaldehyde exposure also causes a range of noncancer health
effects, including irritation of the eyes (burning and watering of the
eyes), nose and throat. Effects from repeated exposure in humans
include respiratory tract irritation, chronic bronchitis and nasal
epithelial lesions such as metaplasia and loss of cilia. Animal studies
suggest that formaldehyde may also cause airway inflammation--including
eosinophil infiltration into the airways. There are several studies
that suggest that formaldehyde may increase the risk of asthma--
particularly in the young.^{373 374}
---------------------------------------------------------------------------

\373\ Agency for Toxic Substances and Disease Registry (ATSDR).
1999. Toxicological profile for Formaldehyde. Atlanta, GA: U.S.
Department of Health and Human Services, Public Health Service.
http://www.atsdr.cdc.gov/toxprofiles/tp111.html Docket EPA-HQ-OAR-
2009-0472-1191.
\374\ WHO (2002) Concise International Chemical Assessment
Document 40: Formaldehyde. Published under the joint sponsorship of
the United Nations Environment Programme, the International Labour
Organization, and the World Health Organization, and produced within
the framework of the Inter-Organization Programme for the Sound
Management of Chemicals. Geneva. Docket EPA-HQ-OAR-2009-0472-1199.
---------------------------------------------------------------------------

iv. Acetaldehyde
Acetaldehyde is classified in EPA's IRIS database as a probable
human carcinogen, based on nasal tumors in rats, and is considered
toxic by the inhalation, oral, and intravenous routes.\375\
Acetaldehyde is reasonably anticipated to be a human carcinogen by the
U.S. DHHS in the 11th Report on Carcinogens and is classified as
possibly carcinogenic to humans (Group 2B) by the
IARC.^{376 377} EPA is currently conducting a reassessment of
cancer risk from inhalation exposure to acetaldehyde.
---------------------------------------------------------------------------

\375\ U.S. EPA. 1991. Integrated Risk Information System File of
Acetaldehyde. Research and Development, National Center for
Environmental Assessment, Washington, DC. This material is available
electronically at http://www.epa.gov/iris/subst/0290.htm. Docket
EPA-HQ-OAR-2009-0472-0390.
\376\ U.S. Department of Health and Human Services National
Toxicology Program 11th Report on Carcinogens available at:
ntp.niehs.nih.gov/index.cfm?objectid=32BA9724-F1F6-975E-
7FCE50709CB4C932.
\377\ International Agency for Research on Cancer (IARC). 1999.
Re-evaluation of some organic chemicals, hydrazine, and hydrogen
peroxide. IARC Monographs on the Evaluation of Carcinogenic Risk of
Chemical to Humans, Vol 71. Lyon, France. Docket EPA-HQ-OAR-2009-
0472-0387.
---------------------------------------------------------------------------

The primary noncancer effects of exposure to acetaldehyde vapors
include irritation of the eyes, skin, and respiratory tract.\378\ In
short-term (4 week) rat studies, degeneration of olfactory epithelium
was observed at various concentration levels of

[[Page 25503]]

acetaldehyde exposure.^{379 380} Data from these studies were
used by EPA to develop an inhalation reference concentration. Some
asthmatics have been shown to be a sensitive subpopulation to
decrements in functional expiratory volume (FEV1 test) and
bronchoconstriction upon acetaldehyde inhalation.\381\ The agency is
currently conducting a reassessment of the health hazards from
inhalation exposure to acetaldehyde.
---------------------------------------------------------------------------

\378\ U.S. EPA. 1991. Integrated Risk Information System File of
Acetaldehyde. This material is available electronically at http://
www.epa.gov/iris/subst/0290.htm.
\379\ Appleman, L. M., R. A. Woutersen, V. J. Feron, R. N.
Hooftman, and W. R. F. Notten. 1986. Effects of the variable versus
fixed exposure levels on the toxicity of acetaldehyde in rats. J.
Appl. Toxicol. 6: 331-336.
\380\ Appleman, L.M., R.A. Woutersen, and V.J. Feron. 1982.
Inhalation toxicity of acetaldehyde in rats. I. Acute and subacute
studies. Toxicology. 23: 293-297. Docket EPA-HQ-OAR-2009-0472-0392.
\381\ Myou, S.; Fujimura, M.; Nishi K.; Ohka, T.; and Matsuda,
T. 1993. Aerosolized acetaldehyde induces histamine-mediated
bronchoconstriction in asthmatics. Am. Rev. Respir.Dis.148(4 Pt 1):
940-3. Docket EPA-HQ-OAR-2009-0472-0408.
---------------------------------------------------------------------------

v. Acrolein
Acrolein is extremely acrid and irritating to humans when inhaled,
with acute exposure resulting in upper respiratory tract irritation,
mucus hypersecretion and congestion. The intense irritancy of this
carbonyl has been demonstrated during controlled tests in human
subjects, who suffer intolerable eye and nasal mucosal sensory
reactions within minutes of exposure.\382\ These data and additional
studies regarding acute effects of human exposure to acrolein are
summarized in EPA's 2003 IRIS Human Health Assessment for
acrolein.\383\ Evidence available from studies in humans indicate that
levels as low as 0.09 ppm (0.21 mg/m\3\) for five minutes may elicit
subjective complaints of eye irritation with increasing concentrations
leading to more extensive eye, nose and respiratory symptoms.\384\
Lesions to the lungs and upper respiratory tract of rats, rabbits, and
hamsters have been observed after subchronic exposure to acrolein.\385\
Acute exposure effects in animal studies report bronchial hyper-
responsiveness.\386\ In a recent study, the acute respiratory irritant
effects of exposure to 1.1 ppm acrolein were more pronounced in mice
with allergic airway disease by comparison to non-diseased mice which
also showed decreases in respiratory rate.\387\ Based on these animal
data and demonstration of similar effects in humans (e.g., reduction in
respiratory rate), individuals with compromised respiratory function
(e.g., emphysema, asthma) are expected to be at increased risk of
developing adverse responses to strong respiratory irritants such as
acrolein.
---------------------------------------------------------------------------

\382\ Sim VM, Pattle RE. Effect of possible smog irritants on
human subjects JAMA165: 1980-2010, 1957. Docket EPA-HQ-OAR-2009-
0472-0395.
\383\ U.S. EPA (U.S. Environmental Protection Agency). (2003)
Toxicological review of acrolein in support of summary information
on Integrated Risk Information System (IRIS) National Center for
Environmental Assessment, Washington, DC. EPA/635/R-03/003.
Available online at: http://www.epa.gov/ncea/iris.
\384\ Weber-Tschopp, A; Fischer, T; Gierer, R; et al. (1977)
Experimentelle reizwirkungen von Acrolein auf den Menschen. Int Arch
Occup Environ Hlth 40(2):117-130. In German Docket EPA-HQ-OAR-2009-
0472-0394.
\385\ Integrated Risk Information System File of Acrolein.
Office of Research and Development, National Center for
Environmental Assessment, Washington, DC. This material is available
at http://www.epa.gov/iris/subst/0364.htm. Docket EPA-HQ-OAR-2009-
0472-0391.
\386\ U.S. EPA (U.S. Environmental Protection Agency). (2003)
Toxicological review of acrolein in support of summary information
on Integrated Risk Information System (IRIS) National Center for
Environmental Assessment, Washington, DC. EPA/635/R-03/003.
Available online at: http://www.epa.gov/ncea/iris.
\387\ Morris JB, Symanowicz PT, Olsen JE, et al. 2003. Immediate
sensory nerve-mediated respiratory responses to irritants in healthy
and allergic airway-diseased mice. J Appl Physiol 94(4):1563-1571.
Docket EPA-HQ-OAR-2009-0472-0396.
---------------------------------------------------------------------------

EPA determined in 2003 that the human carcinogenic potential of
acrolein could not be determined because the available data were
inadequate. No information was available on the carcinogenic effects of
acrolein in humans and the animal data provided inadequate evidence of
carcinogenicity.\388\ The IARC determined in 1995 that acrolein was not
classifiable as to its carcinogenicity in humans.\389\
---------------------------------------------------------------------------

\388\ U.S. EPA 2003. Integrated Risk Information System File of
Acrolein. Research and Development, National Center for
Environmental Assessment, Washington, DC. This material is available
at http://www.epa.gov/iris/subst/0364.htm.
\389\ International Agency for Research on Cancer (IARC). 1995.
Monographs on the evaluation of carcinogenic risk of chemicals to
humans, Volume 63. Dry cleaning, some chlorinated solvents and other
industrial chemicals, World Health Organization, Lyon, France.
Docket EPA-HQ-OAR-2009-0472-0393.
---------------------------------------------------------------------------

vi. Polycyclic Organic Matter (POM)
POM is generally defined as a large class of organic compounds
which have multiple benzene rings and a boiling point greater than 100
degrees Celsius. Many of the compounds included in the class of
compounds known as POM are classified by EPA as probable human
carcinogens based on animal data. One of these compounds, naphthalene,
is discussed separately below. Polycyclic aromatic hydrocarbons (PAHs)
are a subset of POM that contain only hydrogen and carbon atoms. A
number of PAHs are known or suspected carcinogens. Recent studies have
found that maternal exposures to PAHs (a subclass of POM) in a
population of pregnant women were associated with several adverse birth
outcomes, including low birth weight and reduced length at birth, as
well as impaired cognitive development at age three.^{390 391}
EPA has not yet evaluated these recent studies.
---------------------------------------------------------------------------

\390\ Perera, F.P.; Rauh, V.; Tsai, W-Y.; et al. (2002) Effect
of transplacental exposure to environmental pollutants on birth
outcomes in a multiethnic population. Environ Health Perspect. 111:
201-205. Docket EPA-HQ-OAR-2009-0472-0372.
\391\ Perera, F.P.; Rauh, V.; Whyatt, R.M.; Tsai, W.Y.; Tang,
D.; Diaz, D.; Hoepner, L.; Barr, D.; Tu, Y.H.; Camann, D.; Kinney,
P. (2006) Effect of prenatal exposure to airborne polycyclic
aromatic hydrocarbons on neurodevelopment in the first 3 years of
life among inner-city children. Environ Health Perspect 114: 1287-
1292. Docket EPA-HQ-OAR-2009-0472-0373.
---------------------------------------------------------------------------

vii. Naphthalene
Naphthalene is found in small quantities in gasoline and diesel
fuels. Naphthalene emissions have been measured in larger quantities in
both gasoline and diesel exhaust compared with evaporative emissions
from mobile sources, indicating it is primarily a product of
combustion. EPA released an external review draft of a reassessment of
the inhalation carcinogenicity of naphthalene based on a number of
recent animal carcinogenicity studies.\392\ The draft reassessment
completed external peer review.\393\ Based on external peer review
comments received, additional analyses are being undertaken. This
external review draft does not represent official agency opinion and
was released solely for the purposes of external peer review and public
comment. The National Toxicology Program listed naphthalene as
``reasonably anticipated to be a human carcinogen'' in 2004 on the
basis of bioassays reporting clear evidence of carcinogenicity in rats
and some evidence of carcinogenicity in mice.\394\ California EPA has
released a new risk assessment for naphthalene, and the

[[Page 25504]]

IARC has reevaluated naphthalene and re-classified it as Group 2B:
possibly carcinogenic to humans.\395\ Naphthalene also causes a number
of chronic non-cancer effects in animals, including abnormal cell
changes and growth in respiratory and nasal tissues.\396\
---------------------------------------------------------------------------

\392\ U.S. EPA 2004. Toxicological Review of Naphthalene
(Reassessment of the Inhalation Cancer Risk), Environmental
Protection Agency, Integrated Risk Information System, Research and
Development, National Center for Environmental Assessment,
Washington, DC. This material is available electronically at http://
www.epa.gov/iris/subst/0436.htm. Docket EPA-HQ-OAR-2009-0472-0272.
\393\ Oak Ridge Institute for Science and Education. (2004).
External Peer Review for the IRIS Reassessment of the Inhalation
Carcinogenicity of Naphthalene. August 2004. http://cfpub.epa.gov/
ncea/cfm/recordisplay.cfm?deid=84403. Docket EPA-HQ-OAR-2009-0472-
0273.
\394\ National Toxicology Program (NTP). (2004). 11th Report on
Carcinogens. Public Health Service, U.S. Department of Health and
Human Services, Research Triangle Park, NC. Available from: http://
ntp-server.niehs.nih.gov.
\395\ International Agency for Research on Cancer (IARC).
(2002). Monographs on the Evaluation of the Carcinogenic Risk of
Chemicals for Humans. Vol. 82. Lyon, France. Docket EPA-HQ-OAR-2009-
0472-0274.
\396\ U.S. EPA. 1998. Toxicological Review of Naphthalene,
Environmental Protection Agency, Integrated Risk Information System,
Research and Development, National Center for Environmental
Assessment, Washington, DC. This material is available
electronically at http://www.epa.gov/iris/subst/0436.htm.
---------------------------------------------------------------------------

viii. Other Air Toxics
In addition to the compounds described above, other compounds in
gaseous hydrocarbon and PM emissions from vehicles will be affected by
this final rule. Mobile source air toxic compounds that would
potentially be impacted include ethylbenzene, propionaldehyde, toluene,
and xylene. Information regarding the health effects of these compounds
can be found in EPA's IRIS database.\397\
---------------------------------------------------------------------------

\397\ U.S. EPA Integrated Risk Information System (IRIS)
database is available at: http://www.epa.gov/iris.
---------------------------------------------------------------------------

f. Exposure and Health Effects Associated With Traffic
Populations who live, work, or attend school near major roads
experience elevated exposure concentrations to a wide range of air
pollutants, as well as higher risks for a number of adverse health
effects. While the previous sections of this preamble have focused on
the health effects associated with individual criteria pollutants or
air toxics, this section discusses the mixture of different exposures
near major roadways, rather than the effects of any single pollutant.
As such, this section emphasizes traffic-related air pollution, in
general, as the relevant indicator of exposure rather than any
particular pollutant.
Concentrations of many traffic-generated air pollutants are
elevated for up to 300-500 meters downwind of roads with high traffic
volumes.\398\ Numerous sources on roads contribute to elevated roadside
concentrations, including exhaust and evaporative emissions, and
resuspension of road dust and tire and brake wear. Concentrations of
several criteria and hazardous air pollutants are elevated near major
roads. Furthermore, different semi-volatile organic compounds and
chemical components of particulate matter, including elemental carbon,
organic material, and trace metals, have been reported at higher
concentrations near major roads.
---------------------------------------------------------------------------

\398\ Zhou, Y.; Levy, J.I. (2007) Factors influencing the
spatial extent of mobile source air pollution impacts: a meta-
analysis. BMC Public Health 7: 89. doi:10.1186/1471-2458-7-89.
---------------------------------------------------------------------------

Populations near major roads experience greater risk of certain
adverse health effects. The Health Effects Institute published a report
on the health effects of traffic-related air pollution.\399\ It
concluded that evidence is ``sufficient to infer the presence of a
causal association'' between traffic exposure and exacerbation of
childhood asthma symptoms. The HEI report also concludes that the
evidence is either ``sufficient'' or ``suggestive but not sufficient''
for a causal association between traffic exposure and new childhood
asthma cases. A review of asthma studies by Salam et al. (2008) reaches
similar conclusions.\400\ The HEI report also concludes that there is
``suggestive'' evidence for pulmonary function deficits associated with
traffic exposure, but concluded that there is ``inadequate and
insufficient'' evidence for causal associations with respiratory health
care utilization, adult-onset asthma, COPD symptoms, and allergy. A
review by Holguin (2008) notes that the effects of traffic on asthma
may be modified by nutrition status, medication use, and genetic
factors.\401\
---------------------------------------------------------------------------

\399\ HEI Panel on the Health Effects of Air Pollution. (2010)
Traffic-related air pollution: a critical review of the literature
on emissions, exposure, and health effects. [Online at http://
www.healtheffects.org].
\400\ Salam, M.T.; Islam, T.; Gilliland, F.D. (2008) Recent
evidence for adverse effects of residential proximity to traffic
sources on asthma. Current Opin Pulm Med 14: 3-8.
\401\ Holguin, F. (2008) Traffic, outdoor air pollution, and
asthma. Immunol Allergy Clinics North Am 28: 577-588.
---------------------------------------------------------------------------

The HEI report also concludes that evidence is ``suggestive'' of a
causal association between traffic exposure and all-cause and
cardiovascular mortality. There is also evidence of an association
between traffic-related air pollutants and cardiovascular effects such
as changes in heart rhythm, heart attack, and cardiovascular disease.
The HEI report characterizes this evidence as ``suggestive'' of a
causal association, and an independent epidemiological literature
review by Adar and Kaufman (2007) concludes that there is ``consistent
evidence'' linking traffic-related pollution and adverse cardiovascular
health outcomes.\402\
---------------------------------------------------------------------------

\402\ Adar, S.D.; Kaufman, J.D. (2007) Cardiovascular disease
and air pollutants: evaluating and improving epidemiological data
implicating traffic exposure. Inhal Toxicol 19: 135-149.
---------------------------------------------------------------------------

Some studies have reported associations between traffic exposure
and other health effects, such as birth outcomes (e.g., low birth
weight) and childhood cancer. The HEI report concludes that there is
currently ``inadequate and insufficient'' evidence for a causal
association between these effects and traffic exposure. A review by
Raaschou-Nielsen and Reynolds (2006) concluded that evidence of an
association between childhood cancer and traffic-related air pollutants
is weak, but noted the inability to draw firm conclusions based on
limited evidence.\403\
---------------------------------------------------------------------------

\403\ Raaschou-Nielsen, O.; Reynolds, P. (2006) Air pollution
and childhood cancer: A review of the epidemiological literature.
Int J Cancer 118: 2920-2929.
---------------------------------------------------------------------------

There is a large population in the U.S. living in close proximity
of major roads. According to the Census Bureau's American Housing
Survey for 2007, approximately 20 million residences in the U.S., 15.6%
of all homes, are located within 300 feet (91 m) of a highway with 4+
lanes, a railroad, or an airport.\404\ Therefore, at current population
of approximately 309 million, assuming that population and housing
similarly distributed, there are over 48 million people in the U.S.
living near such sources. The HEI report also notes that in two North
American cities, Los Angeles and Toronto, over 40% of each city's
population live within 500 meters of a highway or 100 meters of a major
road. It also notes that about 33% of each city's population resides
within 50 meters of major roads. Together, the evidence suggests that a
large U.S. population lives in areas with elevated traffic-related air
pollution.
---------------------------------------------------------------------------

\404\ U.S. Census Bureau (2008) American Housing Survey for the
United States in 2007. Series H-150 (National Data), Table 1A-6.
[Accessed at http://www.census.gov/hhes/www/housing/ahs/ahs07/
ahs07.html on January 22, 2009]
---------------------------------------------------------------------------

People living near roads are often socioeconomically disadvantaged.
According to the 2007 American Housing Survey, a renter-occupied
property is over twice as likely as an owner-occupied property to be
located near a highway with 4+ lanes, railroad or airport. In the same
survey, the median household income of rental housing occupants was
less than half that of owner-occupants ($28,921/$59,886). Numerous
studies in individual urban areas report higher levels of traffic-
related air pollutants in areas with high minority or poor
populations.^{405 406 407}
---------------------------------------------------------------------------

\405\ Lena, T.S.; Ochieng, V.; Carter, M.; Holgu[iacute]n-Veras,
J.; Kinney, P.L. (2002) Elemental carbon and PM2.5 levels
in an urban community heavily impacted by truck traffic. Environ
Health Perspect 110: 1009-1015.
\406\ Wier, M.; Sciammas, C.; Seto, E.; Bhatia, R.; Rivard, T.
(2009) Health, traffic, and environmental justice: collaborative
research and community action in San Francisco, California. Am J
Public Health 99: S499-S504.
\407\ Forkenbrock, D.J. and L.A. Schweitzer, Environmental
Justice and Transportation Investment Policy. Iowa City: University
of Iowa, 1997.

---------------------------------------------------------------------------

[[Page 25505]]

Students may also be exposed in situations where schools are
located near major roads. In a study of nine metropolitan areas across
the U.S., Appatova et al. (2008) found that on average greater than 33%
of schools were located within 400 m of an Interstate, U.S., or State
highway, while 12% were located within 100 m.\408\ The study also found
that among the metropolitan areas studied, schools in the Eastern U.S.
were more often sited near major roadways than schools in the Western
U.S.
---------------------------------------------------------------------------

\408\ Appatova, A.S.; Ryan, P.H.; LeMasters, G.K.; Grinshpun,
S.A. (2008) Proximal exposure of public schools and students to
major roadways: a nationwide U.S. survey. J Environ Plan Mgmt
---------------------------------------------------------------------------

Demographic studies of students in schools near major roadways
suggest that this population is more likely than the general student
population to be of non-white race or Hispanic ethnicity, and more
often live in low socioeconomic status locations.^{409 410 411}
There is some inconsistency in the evidence, which may be due to
different local development patterns and measures of traffic and
geographic scale used in the studies.\408\
---------------------------------------------------------------------------

\409\ Green, R.S.; Smorodinsky, S.; Kim, J.J.; McLaughlin, R.;
Ostro, B. (2004) Proximity of California public schools to busy
roads. Environ Health Perspect 112: 61-66.
\410\ Houston, D.; Ong, P.; Wu, J.; Winer, A. (2006) Proximity
of licensed child care facilities to near-roadway vehicle pollution.
Am J Public Health 96: 1611-1617.
\411\ Wu, Y.; Batterman, S. (2006) Proximity of schools in
Detroit, Michigan to automobile and truck traffic. J Exposure Sci
Environ Epidemiol 16: 457-470.
---------------------------------------------------------------------------

4. Environmental Effects of Non-GHG Pollutants
In this section we discuss some of the environmental effects of PM
and its precursors such as visibility impairment, atmospheric
deposition, and materials damage and soiling, as well as environmental
effects associated with the presence of ozone in the ambient air, such
as impacts on plants, including trees, agronomic crops and urban
ornamentals, and environmental effects associated with air toxics. No
substantive comments were received on the environmental effects of non-
GHG pollutants.
a. Visibility
Visibility can be defined as the degree to which the atmosphere is
transparent to visible light.\412\ Visibility impairment is caused by
light scattering and absorption by suspended particles and gases.
Visibility is important because it has direct significance to people's
enjoyment of daily activities in all parts of the country. Individuals
value good visibility for the well-being it provides them directly,
where they live and work, and in places where they enjoy recreational
opportunities. Visibility is also highly valued in significant natural
areas, such as national parks and wilderness areas, and special
emphasis is given to protecting visibility in these areas. For more
information on visibility see the final 2009 PM ISA.\413\
---------------------------------------------------------------------------

\412\ National Research Council, 1993. Protecting Visibility in
National Parks and Wilderness Areas. National Academy of Sciences
Committee on Haze in National Parks and Wilderness Areas. National
Academy Press, Washington, DC. Docket EPA-HQ-OAR-2005-0161. This
book can be viewed on the National Academy Press Web site at http://
www.nap.edu/books/0309048443/html/.
\413\ U.S. EPA (2009). Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-08/139F, 2009. Docket EPA-HQ-OAR-
2009-0472-11295.
---------------------------------------------------------------------------

EPA is pursuing a two-part strategy to address visibility. First,
EPA has concluded that PM2.5 causes adverse effects on
visibility in various locations, depending on PM concentrations and
factors such as chemical composition and average relative humidity, and
has set secondary PM2.5 standards.\414\ The secondary
PM2.5 standards act in conjunction with the regional haze
program. The regional haze rule (64 FR 35714) was put in place in July
1999 to protect the visibility in mandatory class I Federal areas.
There are 156 national parks, forests and wilderness areas categorized
as mandatory class I Federal areas (62 FR 38680-81, July 18,
1997).\415\ Visibility can be said to be impaired in both
PM2.5 nonattainment areas and mandatory class I Federal
areas.
---------------------------------------------------------------------------

\414\ The existing annual primary and secondary PM2.5
standards have been remanded and are being addressed in the
currently ongoing PM NAAQS review.
\415\ These areas are defined in CAA section 162 as those
national parks exceeding 6,000 acres, wilderness areas and memorial
parks exceeding 5,000 acres, and all international parks which were
in existence on August 7, 1977.
---------------------------------------------------------------------------

b. Plant and Ecosystem Effects of Ozone
Elevated ozone levels contribute to environmental effects, with
impacts to plants and ecosystems being of most concern. Ozone can
produce both acute and chronic injury in sensitive species depending on
the concentration level and the duration of the exposure. Ozone effects
also tend to accumulate over the growing season of the plant, so that
even low concentrations experienced for a longer duration have the
potential to create chronic stress on vegetation. Ozone damage to
plants includes visible injury to leaves and impaired photosynthesis,
both of which can lead to reduced plant growth and reproduction,
resulting in reduced crop yields, forestry production, and use of
sensitive ornamentals in landscaping. In addition, the impairment of
photosynthesis, the process by which the plant makes carbohydrates (its
source of energy and food), can lead to a subsequent reduction in root
growth and carbohydrate storage below ground, resulting in other, more
subtle plant and ecosystems impacts.
These latter impacts include increased susceptibility of plants to
insect attack, disease, harsh weather, interspecies competition and
overall decreased plant vigor. The adverse effects of ozone on forest
and other natural vegetation can potentially lead to species shifts and
loss from the affected ecosystems, resulting in a loss or reduction in
associated ecosystem goods and services. Lastly, visible ozone injury
to leaves can result in a loss of aesthetic value in areas of special
scenic significance like national parks and wilderness areas. The final
2006 Ozone Air Quality Criteria Document presents more detailed
information on ozone effects on vegetation and ecosystems.
c. Atmospheric Deposition
Wet and dry deposition of ambient particulate matter delivers a
complex mixture of metals (e.g., mercury, zinc, lead, nickel, aluminum,
cadmium), organic compounds (e.g., POM, dioxins, furans) and inorganic
compounds (e.g., nitrate, sulfate) to terrestrial and aquatic
ecosystems. The chemical form of the compounds deposited depends on a
variety of factors including ambient conditions (e.g., temperature,
humidity, oxidant levels) and the sources of the material. Chemical and
physical transformations of the compounds occur in the atmosphere as
well as the media onto which they deposit. These transformations in
turn influence the fate, bioavailability and potential toxicity of
these compounds. Atmospheric deposition has been identified as a key
component of the environmental and human health hazard posed by several
pollutants including mercury, dioxin and PCBs.\416\
---------------------------------------------------------------------------

\416\ U.S. EPA (2000) Deposition of Air Pollutants to the Great
Waters: Third Report to Congress. Office of Air Quality Planning and
Standards. EPA-453/R-00-0005. Docket EPA-HQ-OAR-2009-0472-0091.
---------------------------------------------------------------------------

Adverse impacts on water quality can occur when atmospheric
contaminants deposit to the water surface or when

[[Page 25506]]

material deposited on the land enters a waterbody through runoff.
Potential impacts of atmospheric deposition to waterbodies include
those related to both nutrient and toxic inputs. Adverse effects to
human health and welfare can occur from the addition of excess nitrogen
via atmospheric deposition. The nitrogen-nutrient enrichment
contributes to toxic algae blooms and zones of depleted oxygen, which
can lead to fish kills, frequently in coastal waters. Deposition of
heavy metals or other toxics may lead to the human ingestion of
contaminated fish, impairment of drinking water, damage to the marine
ecology, and limits to recreational uses. Several studies have been
conducted in U.S. coastal waters and in the Great Lakes Region in which
the role of ambient PM deposition and runoff is
investigated.^{417 418 419 420 421}
---------------------------------------------------------------------------

\417\ U.S. EPA (2004) National Coastal Condition Report II.
Office of Research and Development/Office of Water. EPA-620/R-03/
002. Docket EPA-HQ-OAR-2009-0472-0089.
\418\ Gao, Y., E.D. Nelson, M.P. Field, et al. 2002.
Characterization of atmospheric trace elements on PM2.5
particulate matter over the New York-New Jersey harbor estuary.
Atmos. Environ. 36: 1077-1086. Docket EPA-HQ-OAR-2009-0472-11297.
\419\ Kim, G., N. Hussain, J.R. Scudlark, and T.M. Church. 2000.
Factors influencing the atmospheric depositional fluxes of stable
Pb, 210Pb, and 7Be into Chesapeake Bay. J. Atmos. Chem. 36: 65-79.
Docket EPA-HQ-OAR-2009-0472-11299.
\420\ Lu, R., R.P. Turco, K. Stolzenbach, et al. 2003. Dry
deposition of airborne trace metals on the Los Angeles Basin and
adjacent coastal waters. J. Geophys. Res. 108(D2, 4074): AAC 11-1 to
11-24. Docket EPA-HQ-OAR-2009-0472-11296.
\421\ Marvin, C.H., M.N. Charlton, E.J. Reiner, et al. 2002.
Surficial sediment contamination in Lakes Erie and Ontario: A
comparative analysis. J. Great Lakes Res. 28(3): 437-450. Docket
EPA-HQ-OAR-2009-0472-11300.
---------------------------------------------------------------------------

Atmospheric deposition of nitrogen and sulfur contributes to
acidification, altering biogeochemistry and affecting animal and plant
life in terrestrial and aquatic ecosystems across the U.S. The
sensitivity of terrestrial and aquatic ecosystems to acidification from
nitrogen and sulfur deposition is predominantly governed by geology.
Prolonged exposure to excess nitrogen and sulfur deposition in
sensitive areas acidifies lakes, rivers and soils. Increased acidity in
surface waters creates inhospitable conditions for biota and affects
the abundance and nutritional value of preferred prey species,
threatening biodiversity and ecosystem function. Over time, acidifying
deposition also removes essential nutrients from forest soils,
depleting the capacity of soils to neutralize future acid loadings and
negatively affecting forest sustainability. Major effects include a
decline in sensitive forest tree species, such as red spruce (Picea
rubens) and sugar maple (Acer saccharum), and a loss of biodiversity of
fishes, zooplankton, and macro invertebrates.
In addition to the role nitrogen deposition plays in acidification,
nitrogen deposition also leads to nutrient enrichment and altered
biogeochemical cycling. In aquatic systems increased nitrogen can alter
species assemblages and cause eutrophication. In terrestrial systems
nitrogen loading can lead to loss of nitrogen sensitive lichen species,
decreased biodiversity of grasslands, meadows and other sensitive
habitats, and increased potential for invasive species. For a broader
explanation of the topics treated here, refer to the description in
Section 7.1.2 of the RIA.
Adverse impacts on soil chemistry and plant life have been observed
for areas heavily influenced by atmospheric deposition of nutrients,
metals and acid species, resulting in species shifts, loss of
biodiversity, forest decline and damage to forest productivity.
Potential impacts also include adverse effects to human health through
ingestion of contaminated vegetation or livestock (as in the case for
dioxin deposition), reduction in crop yield, and limited use of land
due to contamination.
Atmospheric deposition of pollutants can reduce the aesthetic
appeal of buildings and culturally important articles through soiling,
and can contribute directly (or in conjunction with other pollutants)
to structural damage by means of corrosion or erosion. Atmospheric
deposition may affect materials principally by promoting and
accelerating the corrosion of metals, by degrading paints, and by
deteriorating building materials such as concrete and limestone.
Particles contribute to these effects because of their electrolytic,
hygroscopic, and acidic properties, and their ability to adsorb
corrosive gases (principally sulfur dioxide).
d. Environmental Effects of Air Toxics
Fuel combustion emissions contribute to ambient levels of
pollutants that contribute to adverse effects on vegetation. Volatile
organic compounds (VOCs), some of which are considered air toxics, have
long been suspected to play a role in vegetation damage.\422\ In
laboratory experiments, a wide range of tolerance to VOCs has been
observed.\423\ Decreases in harvested seed pod weight have been
reported for the more sensitive plants, and some studies have reported
effects on seed germination, flowering and fruit ripening. Effects of
individual VOCs or their role in conjunction with other stressors
(e.g., acidification, drought, temperature extremes) have not been well
studied. In a recent study of a mixture of VOCs including ethanol and
toluene on herbaceous plants, significant effects on seed production,
leaf water content and photosynthetic efficiency were reported for some
plant species.\424\
---------------------------------------------------------------------------

\422\ U.S. EPA. 1991. Effects of organic chemicals in the
atmosphere on terrestrial plants. EPA/600/3-91/001. Docket EPA-HQ-
OAR-2009-0472-0401.
\423\ Cape JN, ID Leith, J Binnie, J Content, M Donkin, M
Skewes, DN Price AR Brown, AD Sharpe. 2003. Effects of VOCs on
herbaceous plants in an open-top chamber experiment. Environ.
Pollut. 124:341-343. Docket EPA-HQ-OAR-2009-0472-0357.
\424\ Cape JN, ID Leith, J Binnie, J Content, M Donkin, M
Skewes, DN Price AR Brown, AD Sharpe. 2003. Effects of VOCs on
herbaceous plants in an open-top chamber experiment. Environ.
Pollut. 124:341-343. Docket EPA-HQ-OAR-2009-0472-0357.
---------------------------------------------------------------------------

Research suggests an adverse impact of vehicle exhaust on plants,
which has in some cases been attributed to aromatic compounds and in
other cases to nitrogen oxides.^{425 426 427} The impacts of
VOCs on plant reproduction may have long-term implications for
biodiversity and survival of native species near major roadways. Most
of the studies of the impacts of VOCs on vegetation have focused on
short-term exposure and few studies have focused on long-term effects
of VOCs on vegetation and the potential for metabolites of these
compounds to affect herbivores or insects.
---------------------------------------------------------------------------

\425\ Viskari E-L. 2000. Epicuticular wax of Norway spruce
needles as indicator of traffic pollutant deposition. Water, Air,
and Soil Pollut. 121:327-337. Docket EPA-HQ-OAR-2009-0472-1128.
\426\ Ugrekhelidze D, F Korte, G Kvesitadze. 1997. Uptake and
transformation of benzene and toluene by plant leaves. Ecotox.
Environ. Safety 37:24-29. Docket EPA-HQ-OAR-2009-0472-1142.
\427\ Kammerbauer H, H Selinger, R Rommelt, A Ziegler-Jons, D
Knoppik, B Hock. 1987. Toxic components of motor vehicle emissions
for the spruce Picea abies. Environ. Pollut. 48:235-243. Docket EPA-
HQ-OAR-2009-0472-0358.
---------------------------------------------------------------------------

5. Air Quality Impacts of Non-GHG Pollutants
Air quality modeling was performed to assess the impact of the
vehicle standards on criteria and air toxic pollutants. In this
section, we present information on current modeled levels of pollution
as well as projections for 2030, with respect to ambient
PM2.5, ozone, selected air toxics, visibility levels and
nitrogen and sulfur deposition. The air quality modeling results
indicate that the GHG standards have relatively small but measureable
impacts on ambient concentrations of these pollutants. The results are
discussed in more detail below and in Section 7.2 of the RIA. No
substantive

[[Page 25507]]

comments were received on our plans for non-GHG air quality modeling
that were detailed in the proposal for this rule.
We used the Community Multi-scale Air Quality (CMAQ) photochemical
model, version 4.7.1, for our analysis. This version of CMAQ includes a
number of improvements to previous versions of the model. These
improvements are discussed in Section 7.2 of the RIA.
a. Particulate Matter
i. Current Levels
PM2.5 concentrations exceeding the level of the
PM2.5 NAAQS occur in many parts of the country. In 2005, EPA
designated 39 nonattainment areas for the 1997 PM2.5 NAAQS
(70 FR 943, January 5, 2005). These areas are composed of 208 full or
partial counties with a total population exceeding 88 million. The 1997
PM2.5 NAAQS was revised in 2006 and the 2006 24-hour
PM2.5 NAAQS became effective on December 18, 2006. On
October 8, 2009, the EPA issued final nonattainment area designations
for the 2006 24-hour PM2.5 NAAQS (74 FR 58688, November 13,
2009). These designations include 31 areas composed of 120 full or
partial counties with a population of over 70 million. In total, there
are 54 PM2.5 nonattainment areas composed of 243 counties
with a population of almost 102 million people.
ii. Projected Levels Without This Rule
States with PM2.5 nonattainment areas are required to
take action to bring those areas into compliance in the future. Areas
designated as not attaining the 1997 PM2.5 NAAQS will need
to attain the 1997 standards in the 2010 to 2015 time frame, and then
maintain them thereafter. The 2006 24-hour PM2.5
nonattainment areas will be required to attain the 2006 24-hour
PM2.5 NAAQS in the 2014 to 2019 time frame and then be
required to maintain the 2006 24-hour PM2.5 NAAQS
thereafter. The vehicle standards finalized in this action become
effective in 2012 and therefore may be useful to states in attaining or
maintaining the PM2.5 NAAQS.
EPA has already adopted many emission control programs that are
expected to reduce ambient PM2.5 levels and which will
assist in reducing the number of areas that fail to achieve the
PM2.5 NAAQS. Even so, our air quality modeling projects that
in 2030, with all current controls but excluding the impacts of the
vehicle standards adopted here, at least 9 counties with a population
of almost 28 million may not attain the 1997 annual PM2.5
standard of 15 [mu]g/m³ and 26 counties with a population of
over 41 million may not attain the 2006 24-hour PM2.5
standard of 35 [mu]g/m³. These numbers do not account for
those areas that are close to (e.g., within 10 percent of) the
PM2.5 standards. These areas, although not violating the
standards, will also benefit from any reductions in PM2.5
ensuring long-term maintenance of the PM2.5 NAAQS.
iii. Projected Levels With This Rule
Air quality modeling performed for this final rule shows that in
2030 the majority of the modeled counties will see decreases of less
than 0.05 [mu]g/m³ in their annual PM2.5 design
values. The decreases in annual PM2.5 design values that we
see in some counties are likely due to emission reductions related to
lower gasoline production at existing oil refineries; reductions in
direct PM2.5 emissions and PM2.5 precursor
emissions (NOX and SOX) contribute to reductions
in ambient concentrations of both direct PM2.5 and
secondarily-formed PM2.5. The maximum projected decrease in
an annual PM2.5 design value is 0.07 [mu]g/m³ in
Harris County, TX. There are also a few counties that are projected to
see increases of no more than 0.01 [mu]g/m³ in their annual
PM2.5 design values. These small increases in annual
PM2.5 design values are likely related to downstream
emission increases. On a population-weighted basis, the average modeled
2030 annual PM2.5 design value is projected to decrease by
0.01 [mu]g/m³ due to this final rule. Those counties that
are projected to be above the annual PM2.5 standard in 2030
will see slightly larger population-weighted decreases of 0.03 [mu]g/
m³ in their design values due to this final rule.
In addition to looking at annual PM2.5 design values, we
also modeled the impact of the standards on 24-hour PM2.5
design values. Air quality modeling performed for this final rule shows
that in 2030 the majority of the modeled counties will see changes of
between -0.05 [mu]g/m³ and +0.05 [mu]g/m³ in
their 24-hour PM2.5 design values. The decreases in 24-hour
PM2.5 design values that we see in some counties are likely
due to emission reductions related to lower gasoline production at
existing oil refineries; reductions in direct PM2.5
emissions and PM2.5 precursor emissions (NOX and
SOX) contribute to reductions in ambient concentrations of
both direct PM2.5 and secondarily-formed PM2.5.
The maximum projected decrease in a 24-hour PM2.5 design
value is 0.21 [mu]g/m³ in Harris County, TX. There are also
some counties that are projected to see increases of less than 0.05
[mu]g/m³ in their 24-hour PM2.5 design values.
These small increases in 24-hour PM2.5 design values are
likely related to downstream emission increases. On a population-
weighted basis, the average modeled 2030 24-hour PM2.5
design value is projected to decrease by 0.01 [mu]g/m³ due
to this final rule. Those counties that are projected to be above the
24-hour PM2.5 standard in 2030 will see slightly larger
population-weighted decreases of 0.05 [mu]g/m³ in their
design values due to this final rule.
b. Ozone
i. Current Levels
8-hour ozone concentrations exceeding the level of the ozone NAAQS
occur in many parts of the country. In 2008, the EPA amended the ozone
NAAQS (73 FR 16436, March 27, 2008). The final 2008 ozone NAAQS rule
set forth revisions to the previous 1997 NAAQS for ozone to provide
increased protection of public health and welfare. EPA recently
proposed to reconsider the 2008 ozone NAAQS (75 FR 2938, January 19,
2010). Because of the uncertainty the reconsideration proposal creates
regarding the continued applicability of the 2008 ozone NAAQS, EPA has
used its authority to extend by 1 year the deadline for promulgating
designations for those NAAQS (75 FR 2936, January 19, 2010). The new
deadline is March 12, 2011. EPA intends to complete the reconsideration
by August 31, 2010. If EPA establishes new ozone NAAQS as a result of
the reconsideration, they would replace the 2008 ozone NAAQS and
requirements to designate areas and implement the 2008 NAAQS would no
longer apply.
As of January 6, 2010 there are 51 areas designated as
nonattainment for the 1997 8-hour ozone NAAQS, comprising 266 full or
partial counties with a total population of over 122 million people.
These numbers do not include the people living in areas where there is
a future risk of failing to maintain or attain the 1997 8-hour ozone
NAAQS. The numbers above likely underestimate the number of counties
that are not meeting the ozone NAAQS because the nonattainment areas
associated with the more stringent 2008 8-hour ozone NAAQS have not yet
been designated. Table III.G.5-1 provides an estimate, based on 2005-07
air quality data, of the counties with design values greater than the
2008 8-hour ozone NAAQS of 0.075 ppm.

[[Page 25508]]

Table III.G.5-1--Counties with Design Values Greater Than the Ozone
NAAQS
------------------------------------------------------------------------
Number of
counties Population \a\
------------------------------------------------------------------------
1997 Ozone Standard: Counties within 266 122,343,799
the 54 areas currently designated
as nonattainment (as of 1/6/10)....
2008 Ozone Standard: Additional 156 36,678,478
counties that would not meet the
2008 NAAQS (based on 2006-2008 air
quality data) \b\..................
------------------------------------------------------------------------
Total........................... 422 159,022,277
------------------------------------------------------------------------
Notes:
\a\ Population numbers are from 2000 census data.
\b\ Area designations for the 2008 ozone NAAQS have not yet been made.
Nonattainment for the 2008 Ozone NAAQS would be based on three years
of air quality data from later years. Also, the county numbers in this
row include only the counties with monitors violating the 2008 Ozone
NAAQS. The numbers in this table may be an underestimate of the number
of counties and populations that will eventually be included in areas
with multiple counties designated nonattainment.

ii. Projected Levels Without This Rule
States with 8-hour ozone nonattainment areas are required to take
action to bring those areas into compliance in the future. Based on the
final rule designating and classifying 8-hour ozone nonattainment areas
for the 1997 standard (69 FR 23951, April 30, 2004), most 8-hour ozone
nonattainment areas will be required to attain the ozone NAAQS in the
2007 to 2013 time frame and then maintain the NAAQS thereafter. As
noted, EPA is reconsidering the 2008 ozone NAAQS. If EPA promulgates
different ozone NAAQS in 2010 as a result of the reconsideration, these
standards would replace the 2008 ozone NAAQS and there would no longer
be a requirement to designate areas for the 2008 NAAQS. EPA would
designate nonattainment areas for a potential new 2010 primary ozone
NAAQS in 2011. The attainment dates for areas designated nonattainment
for a potential new 2010 primary ozone NAAQS are likely to be in the
2014 to 2031 timeframe, depending on the severity of the problem.\428\
---------------------------------------------------------------------------

\428\ U.S. EPA 2010, Fact Sheet Revisions to Ozone Standards.
http://www.epa.gov/groundlevelozone/pdfs/fs20100106std.pdf.
---------------------------------------------------------------------------

EPA has already adopted many emission control programs that are
expected to reduce ambient ozone levels and assist in reducing the
number of areas that fail to achieve the ozone NAAQS. Even so, our air
quality modeling projects that in 2030, with all current controls but
excluding the impacts of the vehicle standards, up to 16 counties with
a population of almost 35 million may not attain the 2008 ozone
standard of 0.075 ppm (75 ppb). These numbers do not account for those
areas that are close to (e.g., within 10 percent of) the 2008 ozone
standard. These areas, although not violating the standards, will also
be impacted by changes in ozone as they work to ensure long-term
maintenance of the ozone NAAQS.
iii. Projected Levels With This Rule
We do not expect this rule to have a meaningful impact on ozone
concentrations, given the small magnitude of the ozone impacts and the
fact that much of the impact is due to ethanol assumptions that are
independent of this rule. Our modeling projects increases in ozone
design value concentrations in many areas of the country and decreases
in ozone design value concentrations in a few areas. However, the
increases in ozone design values are not due to the standards finalized
in this rule, but are related to our assumptions about the volume of
ethanol that will be blended into gasoline. The ethanol volumes will be
occurring as a result of the recent Renewable Fuel Standards (RFS2)
rule.\429\
---------------------------------------------------------------------------

\429\ EPA 2010, Renewable Fuel Standard Program (RFS2)
Regulatory Impact Analysis. EPA-420-R-10-006, February 2010. Docket
EPA-HQ-OAR-2009-0472-11332. See also 75 FR 14670, March 26, 2010.
---------------------------------------------------------------------------

The ethanol volume assumptions are discussed in the introduction to
Section III.G of this preamble. We attribute decreased fuel consumption
and production from this program to gasoline only, while assuming
constant ethanol volumes in our reference and control cases. Holding
ethanol volumes constant while decreasing gasoline volumes increases
the market share of 10% ethanol (E10) in the control case. However, the
increased E10 market share is projected to occur regardless of this
rule; in the RFS2 analysis we project 100% E10 by 2014. The air quality
impacts of this effect are included in our analyses for the recent RFS2
rule. As the RFS2 analyses indicate, increasing usage of E10 fuels
(when compared with E0 fuels) can increase NOX emissions and
thereby increase ozone concentrations, especially in NOX-
limited areas where relatively small amounts of NOX enable
ozone to form rapidly.\430\
---------------------------------------------------------------------------

\430\ Sections 3.4.2.1.2 and 3.4.3.3 of the Renewable Fuel
Standard Program (RFS2) Regulatory Impact Analysis, EPA-420-R-10-
006, February 2010. Docket EPA-HQ-OAR-2009-0472-11332.
---------------------------------------------------------------------------

The majority of the ozone design value increases are less than 0.1
ppb. The maximum projected increase in an 8-hour ozone design value is
0.25 ppb in Richland County, South Carolina. As mentioned above there
are some areas which see decreases in their ozone design values. The
decreases in ambient ozone concentration are likely due to projected
upstream emissions decreases in NOX and VOCs from reduced
gasoline production. The maximum decrease projected in an 8-hour ozone
design value is 0.22 ppb in Riverside County, California. On a
population-weighted basis, the average modeled 8-hour ozone design
values are projected to increase by 0.01 ppb in 2030 and the design
values for those counties that are projected to be above the 2008 ozone
standard in 2030 will see population-weighted decreases of 0.10 ppb.
c. Air Toxics
i. Current Levels
The majority of Americans continue to be exposed to ambient
concentrations of air toxics at levels which have the potential to
cause adverse health effects.\431\ The levels of air toxics to which
people are exposed vary depending on where people live and work and the
kinds of activities in which they engage, as discussed in detail in
U.S. EPA's most recent Mobile Source Air Toxics Rule.\432\ According to
the National Air Toxic Assessment

[[Page 25509]]

(NATA) for 2002,\433\ mobile sources were responsible for 47 percent of
outdoor toxic emissions, over 50 percent of the cancer risk, and over
80 percent of the noncancer hazard. Benzene is the largest contributor
to cancer risk of all 124 pollutants quantitatively assessed in the
2002 NATA and mobile sources were responsible for 59 percent of benzene
emissions in 2002. Over the years, EPA has implemented a number of
mobile source and fuel controls resulting in VOC reductions, which also
reduce benzene and other air toxic emissions.
---------------------------------------------------------------------------

\431\ U.S. EPA (2009) 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata2002/. Docket EPA-HQ-OAR-2009-0472-
11321.
\432\ U.S. Environmental Protection Agency (2007). Control of
Hazardous Air Pollutants from Mobile Sources; Final Rule. 72 FR
8434, February 26, 2007. Docket EPA-HQ-OAR-2009-0472-0271, 0271.1
and 0271.2.
\433\ U.S. EPA (2009) 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata2002/. Docket EPA-HQ-OAR-2009-0472-
11321.
---------------------------------------------------------------------------

ii. Projected Levels
Our modeling indicates that the GHG standards have relatively
little impact on national average ambient concentrations of the modeled
air toxics. Additional detail on the air toxics results can be found in
Section 7.2.2.3 of the RIA.
d. Nitrogen and Sulfur Deposition
i. Current Levels
Over the past two decades, the EPA has undertaken numerous efforts
to reduce nitrogen and sulfur deposition across the U.S. Analyses of
long-term monitoring data for the U.S. show that deposition of both
nitrogen and sulfur compounds has decreased over the last 17 years
although many areas continue to be negatively impacted by deposition.
Deposition of inorganic nitrogen and sulfur species routinely measured
in the U.S. between 2004 and 2006 were as high as 9.6 kilograms of
nitrogen per hectare per year (kg N/ha/yr) and 21.3 kilograms of sulfur
per hectare per year (kg S/ha/yr). The data show that reductions were
more substantial for sulfur compounds than for nitrogen compounds.
These numbers are generated by the U.S. national monitoring network and
they likely underestimate nitrogen deposition because neither ammonia
nor organic nitrogen is measured. In the eastern U.S., where data are
most abundant, total sulfur deposition decreased by about 44% between
1990 and 2007, while total nitrogen deposition decreased by 25% over
the same time frame.\434\
---------------------------------------------------------------------------

\434\ U.S. EPA. U.S. EPA's 2008 Report on the Environment (Final
Report). U.S. Environmental Protection Agency, Washington, DC, EPA/
600/R-07/045F (NTIS PB2008-112484). Docket EPA-HQ-OAR-2009-0472-
11298. Updated data available online at: http://cfpub.epa.gov/eroe/
index.cfm?fuseaction=detail.viewInd&ch=46&subtop=341&lv=list.listByCh
apter&r=201744.
---------------------------------------------------------------------------

ii. Projected Levels
Our air quality modeling does not show substantial overall
nationwide impacts on the annual total sulfur and nitrogen deposition
occurring across the U.S. as a result of the vehicle standards required
by this rule. For sulfur deposition the vehicle standards will result
in annual percent decreases of 0.5% to more than 2% in locations with
refineries as a result of the lower output from refineries due to less
gasoline usage. These locations include the Texas and Louisiana
portions of the Gulf Coast; the Washington DC area; Chicago, IL;
portions of Oklahoma and northern Texas; Bismarck, North Dakota;
Billings, Montana; Casper, Wyoming; Salt Lake City, Utah; Seattle,
Washington; and San Francisco, Los Angeles, and San Luis Obispo,
California. The remainder of the country will see only minimal changes
in sulfur deposition, ranging from decreases of less than 0.5% to
increases of less than 0.5%. For a map of 2030 sulfur deposition
impacts and additional information on these impacts, see Section
7.2.2.5 of the RIA. The impacts of the vehicle standards on nitrogen
deposition are minimal, ranging from decreases of up to 0.5% to
increases of up to 0.5%.
e. Visibility
i. Current Levels
As mentioned in Section III.G.5.a, millions of people live in
nonattainment areas for the PM2.5 NAAQS. These populations,
as well as large numbers of individuals who travel to these areas, are
likely to experience visibility impairment. In addition, while
visibility trends have improved in mandatory class I Federal areas, the
most recent data show that these areas continue to suffer from
visibility impairment. In summary, visibility impairment is experienced
throughout the U.S., in multi-State regions, urban areas, and remote
mandatory class I Federal areas.
ii. Projected Levels
Air quality modeling conducted for this final rule was used to
project visibility conditions in 138 mandatory class I Federal areas
across the U.S. in 2030. The results show that all the modeled areas
will continue to have annual average deciview levels above background
in 2030.\435\ The results also indicate that the majority of the
modeled mandatory class I Federal areas will see no change in their
visibility, but some mandatory class I Federal areas will see
improvements in visibility due to the vehicle standards and a few
mandatory class I Federal areas will see visibility decreases. The
average visibility at all modeled mandatory class I Federal areas on
the 20% worst days is projected to improve by 0.002 deciviews, or
0.01%, in 2030. Section 7.2.2.6.2 of the RIA contains more detail on
the visibility portion of the air quality modeling.
---------------------------------------------------------------------------

\435\ The level of visibility impairment in an area is based on
the light-extinction coefficient and a unitless visibility index,
called a ``deciview'', which is used in the valuation of visibility.
The deciview metric provides a scale for perceived visual changes
over the entire range of conditions, from clear to hazy. Under many
scenic conditions, the average person can generally perceive a
change of one deciview. The higher the deciview value, the worse the
visibility. Thus, an improvement in visibility is a decrease in
deciview value.
---------------------------------------------------------------------------

H. What are the estimated cost, economic, and other impacts of the
program?

In this section, EPA presents the costs and impacts of EPA's GHG
program. It is important to note that NHTSA's CAFE standards and EPA's
GHG standards will both be in effect, and each will lead to average
fuel economy increases and CO2 emissions reductions. The two
agencies' standards comprise the National Program, and this discussion
of costs and benefits of EPA's GHG standard does not change the fact
that both the CAFE and GHG standards, jointly, are the source of the
benefits and costs of the National Program. These costs and benefits
are appropriately analyzed separately by each agency and should not be
added together.
This section outlines the basis for assessing the benefits and
costs of the GHG standards and provides estimates of these costs and
benefits. Some of these effects are private, meaning that they affect
consumers and producers directly in their sales, purchases, and use of
vehicles. These private effects include the upfront costs of the
technology, fuel savings, and the benefits of additional driving and
reduced refueling. Other costs and benefits affect people outside the
markets for vehicles and their use; these effects are termed external,
because they affect people in ways other than the effect on the market
for and use of new vehicles and are generally not taken into account by
the purchaser of the vehicle. The external effects include the climate
impacts, the effects on non-GHG pollutants, energy security impacts,
and the effects on traffic, accidents, and noise due to additional
driving. The sum of the private and external benefits and costs is the
net social benefits of the program. There is some debate about the

[[Page 25510]]

role of private benefits in assessing the benefits and costs of the
program: If consumers optimize their purchases of fuel economy, with
full information and perfect foresight, in perfectly efficient markets,
it is possible that they have already considered these benefits in
their vehicle purchase decisions. If so, then no net private benefits
would result from the program, because consumers would already buy
vehicles with the amount of fuel economy that is optimal for them;
requiring additional fuel economy would alter both the purchase prices
of new cars and their lifetime streams of operating costs in ways that
will inevitably reduce consumers' well-being. If these conditions do
not hold, then the private benefits and costs would both count toward
the program's benefits. Section III.H.1 discusses this issue more
fully.
The net benefits of EPA's final program consist of the effects of
the program on:
The vehicle program costs (costs of complying with the
vehicle CO2 standards, taking into account FFV credits
through 2015, the temporary lead-time alternative allowance standard
program (TLAASP), full car/truck trading, and the A/C credit program,
and other flexibilities built into the final program),
Fuel savings associated with reduced fuel usage resulting
from the program,
Greenhouse gas emissions,
Other pollutants,
Noise, congestion, accidents,
Energy security impacts,
Reduced refueling events
Increased driving due to the ``rebound'' effect.

EPA also presents the cost-effectiveness of the standards.
The total monetized benefits (excluding fuel savings) under the
program are projected to be $17.5 to $41.8 billion in 2030, using a 3
percent discount rate applied to the valuation of PM2.5-
related premature mortality and depending on the value used for the
social cost of carbon. The total monetized benefits (excluding fuel
savings) under the program are projected to be $17.4 to $41.7 billion
in 2030, using a 7 percent discount rate applied to the valuation of
PM2.5-related premature mortality and depending on the value
used for the social cost of carbon. These benefits are summarized below
in Table III.H.10-2. The costs of the program in 2030 are estimated to
be approximately $15.8 billion for new vehicle technology less $79.8
billion in savings realized by consumers through fewer fuel
expenditures (calculated using pre-tax fuel prices). These costs are
summarized below in Table III.H.10-1. The estimates developed here use
as a baseline for comparison the fuel economy associated with MY 2011
vehicles. To the extent that greater fuel economy improvements than
those assumed to occur under the baseline may have occurred due to
market forces alone (absent the rule), the analysis overestimates
private and social net benefits.
EPA has undertaken an analysis of the economy-wide impacts of the
GHG tailpipe standards as an exploratory exercise that EPA believes
could provide additional insights into the potential impacts of the
program.\436\ These results were not a factor regarding the
appropriateness of the GHG tailpipe standards. It is important to note
that the results of this modeling exercise are dependent on the
assumptions associated with how producers will make fuel economy
improvements and how consumers will respond to increases in higher
vehicle costs and improved vehicle fuel economy as a result of the
program. Section III.H.1 discusses the underlying distinctions and
implications of the role of consumer response in economic impacts.
---------------------------------------------------------------------------

\436\ See Memorandum to Docket, ``Economy-Wide Impacts of
Proposed Greenhouse Gas Tailpipe Standards,'' March 4, 2010. Docket
EPA-HQ-OAR-2009-0472.
---------------------------------------------------------------------------

Further information on these and other aspects of the economic
impacts of our rule are summarized in the following sections and are
presented in more detail in the RIA for this rulemaking.
1. Conceptual Framework for Evaluating Consumer Impacts
For this rule, EPA projects significant private gains to consumers
in three major areas: (1) Reductions in spending on fuel, (2) time
saved due to less refueling, and (3) welfare gains from additional
driving that results from the rebound effect. In combination, these
private savings, mostly from fuel savings, appear to outweigh by a
large margin the costs of the program, even without accounting for
externalities.
Admittedly, these findings pose an economic conundrum. On the one
hand, consumers are expected to gain significantly from the rules, as
the increased cost of fuel efficient cars appears to be far smaller
than the fuel savings. Yet these technologies are readily available;
financially savvy consumers could have sought vehicles with improved
fuel efficiency, and auto makers seeking those customers could have
offered them. Assuming full information, perfect foresight, perfect
competition, and financially rational consumers and producers, standard
economic theory suggests that normal market operations would have
provided the private net gains to consumers, and the only benefits of
the rule would be due to external benefits. If our analysis projects
net private benefits that consumers have not realized in this perfectly
functioning market, then increased fuel economy should be accompanied
by a corresponding loss in consumer welfare. This calculation assumes
that consumers accurately predict and act on all the benefits they will
get from a new vehicle, and that producers market products providing
those benefits. The existence of large private net benefits from this
rule, then, suggests either that the assumptions noted above do not
hold, or that EPA's analysis has missed some factor(s) tied to improved
fuel economy that reduce(s) consumer welfare.
With respect to the latter, EPA believes the costs of the
technologies developed for this rule take into account the cost needed
to ensure that all vehicle qualities (including performance,
reliability, and size) stay constant, except for fuel economy and
vehicle price. As a result, there would need to be some other changed
qualities that would reduce the benefits consumers receive from their
vehicles. Changing circumstances (e.g., increased demand for horsepower
in response to a drop in fuel prices), and any changes in vehicle
attributes that manufacturers elect to make may result in additional
private impacts to vehicle buyers from requiring increased fuel
economy. Most comments generally supported the cost estimates and the
maintenance of vehicle quality, though two comments expressed concern
over unspecified losses to vehicle quality. Even if there is some such
unidentified loss (which, given existing evidence and modeling
capabilities, is very difficult to quantify), EPA believes that under
realistic assumptions, the private gains from the rule, together with
the social gains (in the form of reduction of externalities), will
continue to substantially outweigh the costs.
The central conundrum has been referred to as the Energy Paradox in
this setting (and in several others).\437\ In short, the problem is
that consumers appear not to purchase products that are in their
economic self-interest. There are

[[Page 25511]]

strong theoretical reasons why this might be so: \438\
---------------------------------------------------------------------------

\437\ Jaffe, A.B., and Stavins, R.N. (1994). The Energy Paradox
and the Diffusion of Conservation Technology. Resource and Energy
Economics, 16(2), 91-122. Docket EPA-HQ-OAR-2009-0472-11415.
\438\ For an overview, see id.
---------------------------------------------------------------------------

Consumers might be myopic and hence undervalue the long-
term.
Consumers might lack information or a full appreciation of
information even when it is presented.
Consumers might be especially averse to the short-term
losses associated with the higher prices of energy efficient products
relative to the uncertain future fuel savings, even if the expected
present value of those fuel savings exceeds the cost (the behavioral
phenomenon of ``loss aversion'')
Even if consumers have relevant knowledge, the benefits of
energy-efficient vehicles might not be sufficiently salient to them at
the time of purchase, and the lack of salience might lead consumers to
neglect an attribute that it would be in their economic interest to
consider.
In the case of vehicle fuel efficiency, and perhaps as a
result of one or more of the foregoing factors, consumers may have
relatively few choices to purchase vehicles with greater fuel economy
once other characteristics, such as vehicle class, are chosen.\439\
---------------------------------------------------------------------------

\439\ For instance, the range of fuel economy (combined city and
highway) available among all listed 2010 6-cylinder minivans is 18
to 20 miles per gallon. With a manual-transmission 4-cylinder
minivan, it is possible to get 24 mpg. See http://
www.fueleconomy.gov, which is jointly maintained by the U.S.
Department of Energy and the EPA. For recent but unpublished
evidence, see Allcott, Hunt, and Nathan Wozny, ``Gasoline Prices,
Fuel Economy, and the Energy Paradox'' (2010), available at http://
web.mit.edu/allcott/www/Allcott%20and%20Wozny%202010%20-
%20Gasoline%20Prices,%20Fuel%20Economy,%20and%20the%20Energy%20Parado
x.pdf.
---------------------------------------------------------------------------

A great deal of work in behavioral economics identifies and
elaborates factors of this sort, which help account for the Energy
Paradox.\440\ This point holds in the context of fuel savings (the main
focus here), but it applies equally to the other private benefits,
including reductions in refueling time and additional driving.\441\ For
example, it might well be questioned whether significant reductions in
refueling time, and corresponding private savings, are fully
internalized when consumers are making purchasing decisions.
---------------------------------------------------------------------------

\440\ Jaffe, A.B., and Stavins, R.N. (1994). The Energy Paradox
and the Diffusion of Conservation Technology. Resource and Energy
Economics, 16(2), 91-122. Docket EPA-HQ-OAR-2009-0472-11415. See
also Allcott and Wozny, supra note.
\441\ For example, it might be maintained that, at the time of
purchase, consumers take full account of the time spent refueling
potentially saved by fuel-efficient cars, but it might also be
questioned whether they have adequate information to do so, or
whether that factor is sufficiently salient to play the proper role
in purchasing decisions.
---------------------------------------------------------------------------

Considerable research findings indicate that the Energy Paradox is
real and significant but the literature has not reached a consensus
about the reasons for its existence. Several researchers have found
evidence suggesting that consumers do not give full or appropriate
weight to fuel economy in purchasing decisions. For example, Sanstad
and Howarth \442\ argue that consumers optimize behavior without full
information by resorting to imprecise but convenient rules of thumb.
Some studies find that a substantial portion of this undervaluation can
be explained by inaccurate assessments of energy savings, or by
uncertainty and irreversibility of energy investments due to
fluctuations in energy prices.\443\ For a number of reasons, consumers
may undervalue future energy savings due to routine mistakes in how
they evaluate these trade-offs. For instance, the calculation of fuel
savings is complex, and consumers may not make it correctly.\444\ The
attribute of fuel economy may be insufficiently salient, leading to a
situation in which consumers pay less than $1 for an expected $1
benefit in terms of discounted gasoline costs.\445\ Larrick and Soll
(2008) find that consumers do not understand how to translate changes
in miles-per-gallon into fuel savings (a concern that EPA is continuing
to attempt to address).\446\ In addition, future fuel price (a major
component of fuel savings) is highly uncertain. Consumer fuel savings
also vary across individuals, who travel different amounts and have
different driving styles. Cost calculations based on the average do not
distinguish between those that may gain or lose as a result of the
policy.\447\ Studies regularly show that fuel economy plays a role in
consumers' vehicle purchases, but modeling that role is still in
development, and there is no consensus that most consumers make fully
informed tradeoffs.\448\
---------------------------------------------------------------------------

\442\ Sanstad, A., and R. Howarth (1994). `` `Normal' Markets,
Market Imperfections, and Energy Efficiency.'' Energy Policy 22(10):
811-818 (Docket EPA-HQ-OAR-2009-0472-11416).
\443\ Greene, D., J. German, and M. Delucchi (2009). ``Fuel
Economy: The Case for Market Failure'' in Reducing Climate Impacts
in the Transportation Sector, Sperling, D., and J. Cannon, eds.
Springer Science (Docket EPA-HQ-OAR-2009-0472-11538); Dasgupta, S.,
S. Siddarth, and J. Silva-Risso (2007). ``To Lease or to Buy? A
Structural Model of a Consumer's Vehicle and Contract Choice
Decisions.'' Journal of Marketing Research 44: 490-502 (Docket EPA-
HQ-OAR-2009-0472-11539); Metcalf, G., and D. Rosenthal (1995). ``The
`New' View of Investment Decisions and Public Policy Analysis: An
Application to Green Lights and Cold Refrigerators,'' Journal of
Policy Analysis and Management 14: 517-531 (Docket EPA-HQ-OAR-2009-
0472-11540); Hassett, K., and G. Metcalf (1995), ``Energy Tax
Credits and Residential Conservation Investment: Evidence from Panel
Data,'' Journal of Public Economics 57: 201-217 (Docket EPA-HQ-OAR-
2009-0472-11543); Metcalf, G., and K. Hassett (1999), ``Measuring
the Energy Savings from Home Improvement Investments: Evidence from
Monthly Billing Data,'' The Review of Economics and Statistics
81(3): 516-528 (Docket EPA-HQ-OAR-2009-0472-0051); van Soest D., and
E. Bulte (2001), ``Does the Energy-Efficiency Paradox Exist?
Technological Progress and Uncertainty.'' Environmental and Resource
Economics 18: 101-12 (Docket EPA-HQ-OAR-2009-0472-11542).
\444\ Turrentine, T. and K. Kurani (2007). ``Car Buyers and Fuel
Economy?'' Energy Policy 35: 1213-1223 (Docket EPA-HQ-OAR-2009-
0472); Larrick, R.P., and J.B. Soll (2008). ``The MPG illusion.''
Science 320: 1593-1594 (Docket EPA-HQ-OAR-2009-0472-0041).
\445\ Allcott, Hunt, and Nathan Wozny, ``Gasoline Prices, Fuel
Economy, and the Energy Paradox'' (2010), available at http://
web.mit.edu/allcott/www/Allcott%20and%20Wozny%202010%20-
%20Gasoline%20Prices,%20Fuel%20Economy,%20and%20the%20Energy%20Parado
x.pdf (Docket EPA-HQ-OAR-2009-0472-11554).
\446\ Sanstad, A., and R. Howarth (1994). `` `Normal' Markets,
Market Imperfections, and Energy Efficiency.'' Energy Policy 22(10):
811-818 (Docket EPA-HQ-OAR-2009-0472-11415); Larrick, R. P., and
J.B. Soll (2008). ``The MPG illusion.'' Science 320: 1593-1594
(Docket EPA-HQ-OAR-2009-0472-0043).
\447\ Hausman J., Joskow P. (1982). ``Evaluating the Costs and
Benefits of Appliance Efficiency Standards.'' American Economic
Review 72: 220-25 (Docket EPA-HQ-OAR-2009-0472-11541).
\448\ E.g., Goldberg, Pinelopi Koujianou, ``Product
Differentiation and Oligopoly in International Markets: The Case of
the U.S. Automobile Industry,'' Econometrica 63(4) (July 1995): 891-
951 (Docket EPA-HQ-OAR-2009-0472-0021); Goldberg, Pinelopi
Koujianou, ``The Effects of the Corporate Average Fuel Efficiency
Standards in the U.S.,'' Journal of Industrial Economics 46(1)
(March 1998): 1-33 (Docket EPA-HQ-OAR-2009-0472-0017); Busse, Meghan
R., Christopher R. Knittel, and Florian Zettelmeyer (2009). ``Pain
at the Pump: How Gasoline Prices Affect Automobile Purchasing in New
and Used Markets,'' Working paper (accessed 6/30/09), available at
http://www.econ.ucdavis.edu/faculty/knittel/papers/gaspaper_
latest.pdf. (Docket EPA-HQ-OAR-2009-0472-0044).
---------------------------------------------------------------------------

Some studies find that a substantial portion of the Energy Paradox
can be explained in models of consumer behavior. For instance, one set
of studies finds that accounting for uncertainty in fuel savings over
time due to unanticipated changes in fuel prices goes a long way toward
explaining this paradox. In this case, consumers give up some uncertain
future fuel savings to avoid higher upfront costs.
A recent review commissioned by EPA supports the finding of great
variability, by looking at one key parameter: The role of fuel economy
in consumers' vehicle purchase decisions.\449\ The review finds no

[[Page 25512]]

consensus on the role of fuel economy in consumer purchase decisions.
Of 27 studies, significant numbers of them find that consumers
undervalue, overvalue, or value approximately correctly the fuel
savings that they will receive from improved fuel economy. The
variation in the value of fuel economy in these studies is so high that
it appears to be inappropriate to identify one central estimate from
the literature. Thus, estimating consumer response to higher vehicle
fuel economy is still unsettled science.
---------------------------------------------------------------------------

\449\ Greene, David L. ``How Consumers Value Fuel Economy: A
Literature Review.'' EPA Report EPA-420-R-10-008, March 2010 (Docket
EPA-HQ-OAR-2009-0472-11575).
---------------------------------------------------------------------------

If there is a difference between fuel savings and consumers'
willingness to pay for fuel savings, the next question is, which is the
appropriate measure of consumer benefit? Fuel savings measure the
actual monetary value that consumers will receive after purchasing a
vehicle; the willingness to pay for fuel economy measures the value
that, before a purchase, consumers place on additional fuel economy. As
noted, there are a number of reasons that consumers may incorrectly
estimate the benefits that they get from improved fuel economy,
including risk or loss aversion, and poor ability to calculate savings.
Also as noted, fuel economy may not be as salient as other vehicle
characteristics when a consumer is considering vehicles. If these
arguments are valid, then there will be significant gains to consumers
of the government mandating additional fuel economy.
EPA requested and received a number of comments discussing the role
of the Energy Paradox in consumer vehicle purchase decisions. Ten
commenters, primarily from a number of academic and non-governmental
organizations, argued that there is a gap between the fuel economy that
consumers purchased and the cost-effective amount, due to a number of
market and behavioral phenomena. These include consumers having
inadequate information about future fuel savings relative to up-front
costs; imperfect competition among auto manufacturers; lack of choice
over fuel economy within classes; lack of salience of fuel economy
relative to other vehicle features at the time of vehicle purchase;
consumer use of heuristic decision-making processes or other rules of
thumb, rather than analyzing fuel economy decisions; consumer risk and
loss aversion leading to more attention to up-front costs than future
fuel savings; and consumer emphasis on visible, status-providing
features of vehicles more than on relatively invisible features such as
fuel economy. The RIA, Chapter 8.1.2, includes further discussion of
these phenomena.
Because of the gap between the fuel economy consumers purchase and
the cost-effective amount, those and additional commenters support
using the full value of fuel savings as a benefit of the rule. A few
asserted, in addition, that auto companies would benefit from offering
vehicles with improved fuel economy. Automakers might underprovide fuel
economy because they believe consumers would not buy it, or that it is
not as salient as price when consumers are buying a vehicle. The
commenters who supported the existence of the gap cite these phenomena
as a basis for regulation of fuel economy. In contrast, two commenters
(the United Auto Workers and one nonprofit research organization)
argued that the market for fuel economy works efficiently; consumers
reveal through their purchase decisions that additional fuel economy is
not important for them. These commenters expressed concern that
regulation to promote more fuel economy would limit consumers' choices
as well as the value of the vehicles to consumers. Yet other commenters
(including some states) noted that the rule protects the existing
variety and choice of vehicles in the market; for this reason, the
value of vehicles to consumers should not suffer as a result of the
rule.
While acknowledging the diversity of perspectives, EPA continues to
include the full fuel savings as private benefits of the rule. Improved
fuel economy will significantly reduce consumer expenditures on fuel,
thus benefiting consumers. It is true that limitations in modeling
affect our ability to estimate how much of these savings would have
occurred in the absence of the rule. For example, some of the
technologies predicted to be adopted in response to the rule may
already be developing due to shifts in consumer demand for fuel
economy. It is possible that some of these savings would have occurred
in the absence of the rule. To the extent that greater fuel economy
improvements than those assumed to occur under the baseline may have
occurred due to market forces alone (absent the rule), the analysis
overestimates private and social net benefits. In the absence of robust
means to identify the changes in fuel economy that would have occurred
without the rule, we estimate the benefits and costs under the
assumption that the rule will lead to more fuel-efficient vehicles than
would have occurred without the rule. As discussed below, limitations
in modeling also affect our ability to estimate the effects of the rule
on net benefits in the market for vehicles.
Consumer vehicle choice models estimate what vehicles consumers buy
based on vehicle and consumer characteristics. In principle, such
models could provide a means of understanding both the role of fuel
economy in consumers' purchase decisions and the effects of this rule
on the benefits that consumers will get from vehicles. The NPRM
included a discussion of the wide variation in the structure and
results of these models. Models or model results have not frequently
been systematically compared to each other. When they have, the results
show large variation over, for instance, the value that consumers place
on additional fuel economy. As a result, EPA found that further
assessment needed to be done before adopting a consumer vehicle choice
model. In the NPRM, EPA asked for comment on the state of the art of
consumer vehicle choice modeling and whether it is sufficiently
developed for use in regulatory analysis.
The responses were varied. Of the six commenters on this issue,
five supported EPA's performing consumer vehicle choice modeling, but
only in general terms; they did not provide recommendations for how to
evaluate the quality of different models or identify a model
appropriate for EPA's purposes. One commenter argued that, if key
differences across models were controlled, then different models would
produce similar results, but there were no suggestions for what choices
to make to control the key differences. One commenter specifically
asked for estimates that quantify losses to consumer welfare. Two
commenters mentioned the importance of taking into account any losses
in vehicle attributes due to increasing fuel economy, but without
specific guidance for how to do so. Some commenters, including some who
supported the use of these models, highlighted some of the models'
potential limitations. Two commenters noted the challenges of modeling
for vehicles that are not yet in the market. Most consumer vehicle
choice models are based on existing vehicle fleets. Future vehicles
will present combinations of vehicle characteristics not previously
seen in markets, such as higher fuel economy and higher price with
other characteristics constant; the existing models may not do well in
predicting consumer responses to these changes. One comment suggested
that the models might be sufficient for predicting changes in consumer
purchase patterns, but not for calculating the welfare gains and losses
to consumers of the changes.
EPA has not used a consumer vehicle choice model for the final rule
analysis, due to concerns we explained in the

[[Page 25513]]

proposal (and discussed in Chapter 8.1 of the RIA), and because no new
information became available to resolve those concerns. It is likely
that variation exists in measuring consumer response to changes in fuel
economy as well as other vehicle characteristics, such as performance.
Thus, there does not appear to be evidence at this time to develop
robust estimates of consumer welfare effects of changes in vehicle
attributes. As noted earlier, EPA's and NHTSA's cost estimates are
based on maintaining these other vehicle attributes. Comments generally
supported the finding that our cost and technology estimates succeeded
in maintaining these other attributes.
EPA will continue its efforts to review the literature, but, given
the known difficulties, EPA has not conducted an analysis using these
models for this program. These issues are discussed in detail in RIA
Chapter 8.1.2.
The next issue is the potential for loss in consumer welfare due to
the rule. As mentioned above (and discussed more thoroughly in Section
III.D of this preamble), the technology cost estimates developed here
take into account the costs to hold other vehicle attributes, such as
size and performance, constant. In addition, the analysis assumes that
the full technology costs are passed along to consumers. With these
assumptions, because welfare losses are monetary estimates of how much
consumers would have to be compensated to be made as well off as in the
absence of the change,\450\ the price increase measures the loss to the
consumer.\451\ Assuming that the full technology cost gets passed along
to the consumer as an increase in price, the technology cost thus
measures the welfare loss to the consumer. Increasing fuel economy
would have to lead to other changes in the vehicles that consumers find
undesirable for there to be additional losses not included in the
technology costs.
---------------------------------------------------------------------------

\450\ This approach describes the economic concept of
compensating variation, a payment of money after a change that would
make a consumer as well off after the change as before it. A related
concept, equivalent variation, estimates the income change that
would be an alternative to the change taking place. The difference
between them is whether the consumer's point of reference is her
welfare before the change (compensating variation) or after the
change (equivalent variation). In practice, these two measures are
typically very close together.
\451\ Indeed, it is likely to be an overestimate of the loss to
the consumer, because the consumer has choices other than buying the
same vehicle with a higher price; she could choose a different
vehicle, or decide not to buy a new vehicle. The consumer would
choose one of those options only if the alternative involves less
loss than paying the higher price. Thus, the increase in price that
the consumer faces would be the upper bound of loss of consumer
welfare, unless there are other changes to the vehicle due to the
fuel economy improvements that make the vehicle less desirable to
consumers.
---------------------------------------------------------------------------

At this time EPA has no available methods to estimate potential
additional effects on consumers not included in the technology cost
estimates, e.g., due to changes in vehicles that consumers find
undesirable, shifts in consumer demand for other attributes, and
uncertainties about the long term reliability of new technologies.
Comments on the rule generally supported EPA's analysis of the
technology costs and the assumption that other vehicle characteristics
were not adversely affected. Any consumer welfare loss cannot be
quantified at this time. For reasons stated above, EPA believes that
any such loss is likely far smaller than the private gains, including
fuel savings and reduced refueling time.
Chapter 8.1 of the RIA discusses in more depth the research on the
Energy Paradox and the state of the art of consumer vehicle choice
modeling.
2. Costs Associated With the Vehicle Program
In this section, EPA presents our estimate of the costs associated
with the final vehicle program. The presentation here summarizes the
costs associated with the new vehicle technology expected to be added
to meet the new GHG standards, including hardware costs to comply with
the A/C credit program. The analysis summarized here provides our
estimate of incremental costs on a per vehicle basis and on an annual
total basis.
The presentation here summarizes the outputs of the OMEGA model
that was discussed in some detail in Section III.D of this preamble.
For details behind the analysis such as the OMEGA model inputs and the
estimates of costs associated with individual technologies, the reader
is directed to Chapters 1 and 2 of the RIA, and Chapter 3 of the Joint
TSD. For more detail on the outputs of the OMEGA model and the overall
vehicle program costs summarized here, the reader is directed to
Chapters 4 and 7 of the RIA.
With respect to the cost estimates for vehicle technologies, EPA
notes that, because these estimates relate to technologies which are in
most cases already available, these cost estimates are technically
robust. Some comments were received that addressed the technology costs
that served as inputs to the OMEGA model as was mentioned in Section
II.E. While those comments did not result in changes to the technology
cost inputs, the technology cost estimates for a select group of
technologies have changed since the NPRM thus changing the vehicle
program costs presented here. These changes, as summarized in Section
II.E and in Chapter 3 of the Joint TSD, were made in response to
updated cost estimates, from the FEV teardown study, available to the
agencies shortly after publication of the NPRM, not in response to
comments. Those cost changes are summarized in Section II.E and in
Chapter 3 of the Joint TSD. EPA believes that we have been conservative
in estimating the vehicle hardware costs associated with this rule.
With respect to the aggregate cost estimations presented in Section
III.H.2.b, EPA notes that there are a number of areas where the results
of our analysis may be conservative and, in general, EPA believes we
have directionally overestimated the costs of compliance with these new
standards, especially in not accounting for the full range of credit
opportunities available to manufacturers. For example, some cost saving
programs are considered in our analysis, such as full car/truck
trading, while others are not, such as early credit generation and
advanced vehicle technology credits.
a. Vehicle Compliance Costs Associated With the CO2
Standards
For the technology and vehicle package costs associated with adding
new CO2-reducing technology to vehicles, EPA began with
EPA's 2008 Staff Report and NHTSA's 2011 CAFE FRM both of which
presented costs generated using existing literature, meetings with
manufacturers and parts suppliers, and meetings with other experts in
the field of automotive cost estimation.\452\ EPA has updated some of
those technology costs with new information from our contract with FEV,
through further discussion with NHTSA, and by converting from 2006
dollars to 2007 dollars using the GDP price deflator. The estimated
costs presented here represent the incremental costs associated with
this rule relative to what the future vehicle fleet would be expected
to look like absent this rule. A more detailed description of the
factors considered in our reference case is presented in Section III.D.
---------------------------------------------------------------------------

\452\ ``EPA Staff Technical Report: Cost and Effectiveness
Estimates of Technologies Used to Reduce Light-Duty Vehicle Carbon
Dioxide Emissions,'' EPA 420-R-08-008; NHTSA 2011 CAFE FRM is at 74
FR 14196; both documents are contained in Docket EPA-HQ-OAR-2009-
0472.
---------------------------------------------------------------------------

The estimates of vehicle compliance costs cover the years of
implementation of the program--2012 through 2016. EPA has also
estimated compliance costs for the years following implementation so
that we can shed

[[Page 25514]]

light on the long term (2022 and later) cost impacts of the
program.\453\ EPA used the year 2022 here because our short-term and
long-term markup factors described shortly below are applied in five
year increments with the 2012 through 2016 implementation span and the
2017 through 2021 span both representing the short-term. Some of the
individual technology cost estimates are presented in brief in Section
III.D, and account for both the direct and indirect costs incurred in
the automobile manufacturing and dealer industries (for a complete
presentation of technology costs, please refer to Chapter 3 of the
Joint TSD). To account for the indirect costs, EPA has applied an
indirect cost markup (ICM) factor to all of our direct costs to arrive
at the estimated technology cost.\454\ The ICM factors used range from
1.11 to 1.64 in the short-term (2012 through 2021), depending on the
complexity of the given technology, to account for differences in the
levels of R&D, tooling, and other indirect costs that will be incurred.
Once the program has been fully implemented, some of the indirect costs
will no longer be attributable to these standards and, as such, a lower
ICM factor is applied to direct costs in years following full
implementation. The ICM factors used range from 1.07 to 1.39 in the
long-term (2022 and later) depending on the complexity of the given
technology.\455\ Note that the short-term ICMs are used in the 2012
through 2016 years of implementation and continue through 2021. EPA
does this since the standards are still being implemented during the
2012 through 2016 model years. Therefore, EPA considers the five year
period following full implementation also to be short-term. Note that,
in general the comments received were supportive of our use of ICMs as
opposed to the more traditional Retail Price Equivalent (RPE).\456\
However, we did receive some comment that we applied inappropriate ICM
factors to some technologies. We have not changed our approach in
response to those comments as explained in greater detail in our
Response to Comments document.
---------------------------------------------------------------------------

\453\ Note that the assumption made here is that the standards
would continue to apply for years beyond 2016 so that new vehicles
sold in model years 2017 and later would continue to incur costs as
a result of this rule. Those costs are estimated to get lower in
2022 because some of the indirect costs attributable to this rule in
the years prior to 2022 would be eliminated in 2022 and later.
\454\ Need to add the recent reference for this study by RTI.
Alex Rogozhin et al., Automobile Industry Regail Price Equivalent
and Indirect Cost Multipliers. Prepared for EPA by RTI International
and Transportation Research Institute, University of Michigan. EPA-
420-R-09-003, February 2009 (Docket EPA-HQ-OAR-2009-0472).
\455\ Gloria Helfand and Todd Sherwood, ``Documentation of the
Development of Indirect Cost Multipliers for Three Automotive
Technologies,'' Office of Transportation and Air Quality, U.S. EPA,
August 2009 (Docket EPA-HQ-OAR-2009-0472).
\456\ The RPE is based on the historical relationship between
direct costs and consumer prices; it is intended to reflect the
average markup over time required to sustain the industry as a
viable operation. Unlike the RPE approach, the ICM focuses more
narrowly on the changes that are required in direct response to
regulation-induced vehicle design changes which may not directly
influence all of the indirect costs that are incurred in the normal
course of business. For example, an RPE markup captures all indirect
costs including costs such as the retirement benefits of retired
employees. However, the retirement benefits for retired employees
are not expected to change as a result of a new GHG regulation and,
therefore, those indirect costs should not increase in relation to
newly added hardware in response to a regulation.
---------------------------------------------------------------------------

EPA has also considered the impacts of manufacturer learning on the
technology cost estimates. Consistent with past EPA rulemakings, EPA
has estimated that some costs would decline by 20 percent with each of
the first two doublings of production beginning with the first year of
implementation. These volume-based cost declines, which EPA calls
``volume'' based learning, take place after manufacturers have had the
opportunity to find ways to improve upon their manufacturing processes
or otherwise manufacture these technologies in a more efficient way.
After two 20 percent cost reduction steps, the cost reduction learning
curve flattens out considerably as only minor improvements in
manufacturing techniques and efficiencies remain to be had. By then,
costs decline roughly three percent per year as manufacturers and
suppliers continually strive to reduce costs. These time-based cost
declines, which EPA calls ``time'' based learning, take place at a rate
of three percent per year. EPA has considered learning impacts on most
but not all of the technologies expected to be used because some of the
expected technologies are already used rather widely in the industry
and, presumably, learning impacts have already occurred. EPA has
considered volume-based learning for only a handful of technologies
that EPA considers to be new or emerging technologies such as the
hybrids and electric vehicles. For most technologies, EPA has
considered them to be more established given their current use in the
fleet and, hence, we have applied the lower time based learning. We
have more discussion of our learning approach and the technologies to
which we have applied which type of learning in Chapter 3 of the Joint
TSD.
The technology cost estimates discussed in Section III.D and
detailed in Chapter 3 of the Joint TSD are used to build up technology
package cost estimates which are then used as inputs to the OMEGA
model. EPA discusses our technology packages and package costs in
Chapter 1 of the RIA. The model determines what level of CO2
improvement is required considering the reference case for each
manufacturer's fleet. The vehicle compliance costs are the outputs of
the model and take into account FFV credits through 2015, TLAAS, full
car/truck trading, and the A/C credit program. Table III.H.2-1 presents
the fleet average incremental vehicle compliance costs for this rule.
As the table indicates, 2012-2016 costs increase every year as the
standards become more stringent. Costs per car and per truck then
remain stable through 2021 while cost per vehicle (car/truck combined)
decline slightly as the fleet mix trends slowly to increasing car
sales. In 2022, costs per car and per truck decline as the long-term
ICM is applied because some indirect costs decrease or are no longer
considered attributable to the program (e.g., warranty costs go down).
Costs per car and per truck remain constant thereafter while the cost
per vehicle declines slightly as the fleet continues to trend toward
cars. By 2030, projections of fleet mix changes become static and the
cost per vehicle remains constant. EPA has a more detailed presentation
of vehicle compliance costs on a manufacturer by manufacturer basis in
Chapter 6 of the RIA.

Table III.H.2-1--Industry Average Vehicle Compliance Costs Associated
With the Tailpipe CO2 Standards
[$/vehicle in 2007 dollars]
------------------------------------------------------------------------
$/vehicle
(car &
Calendar year $/car $/truck truck
combined)
------------------------------------------------------------------------
2012............................. $342 $314 $331

[[Page 25515]]

2013............................. 507 496 503
2014............................. 631 652 639
2015............................. 749 820 774
2016............................. 869 1,098 948
2017............................. 869 1,098 947
2018............................. 869 1,098 945
2019............................. 869 1,098 943
2020............................. 869 1,098 940
2021............................. 869 1,098 939
2022............................. 817 1,032 882
2030............................. 817 1,032 878
2040............................. 817 1,032 875
2050............................. 817 1,032 875
------------------------------------------------------------------------

b. Annual Costs of the Vehicle Program
The costs presented here represent the incremental costs for newly
added technology to comply with the final program. Together with the
projected increases in car and light-truck sales, the increases in per-
vehicle average costs shown in Table III.H.2-1 above result in the
total annual costs reported in Table III.H.2-2 below. Note that the
costs presented in Table III.H.2-2 do not include the savings that
would occur as a result of the improvements to fuel consumption. Those
impacts are presented in Section III.H.4.

Table III.H.2-2--Quantified Annual Costs Associated With the Vehicle
Program
[$Millions of 2007 dollars]
------------------------------------------------------------------------
Quantified
Year annual costs
------------------------------------------------------------------------
2012.................................................... $4,900
2013.................................................... 8,000
2014.................................................... 10,300
2015.................................................... 12,700
2016.................................................... 15,600
2020.................................................... 15,600
2030.................................................... 15,800
2040.................................................... 17,400
2050.................................................... 19,000
NPV, 3%................................................. 345,900
NPV, 7%................................................. 191,900
------------------------------------------------------------------------

3. Cost per Ton of Emissions Reduced
EPA has calculated the cost per ton of GHG (CO2-
equivalent, or CO2e) reductions associated with this rule
using the above costs and the emissions reductions described in Section
III.F. More detail on the costs, emission reductions, and the cost per
ton can be found in the RIA and Joint TSD. EPA has calculated the cost
per metric ton of GHG emissions reductions in the years 2020, 2030,
2040, and 2050 using the annual vehicle compliance costs and emission
reductions for each of those years. The value in 2050 represents the
long-term cost per ton of the emissions reduced. EPA has also
calculated the cost per metric ton of GHG emission reductions including
the savings associated with reduced fuel consumption (presented below
in Section III.H.4). This latter calculation does not include the other
benefits associated with this rule such as those associated with
criteria pollutant reductions or energy security benefits as discussed
later in sections III.H.4 through III.H.9. By including the fuel
savings in the cost estimates, the cost per ton is less than $0, since
the estimated value of fuel savings outweighs the vehicle program
costs. With regard to the CH4 and N2O standards,
since these standards will be emissions caps designed to ensure that
manufacturers do not backslide from current levels, EPA has not
estimated costs associated with the standards (since the standards will
not require any change from current practices nor does EPA estimate
they will result in emissions reductions).
The results for CO2e costs per ton under the rule are
shown in Table III.H.3-1.

Table III.H.3-1-- Annual Cost Per Metric Ton of CO2e Reduced, in $2007 Dollars
----------------------------------------------------------------------------------------------------------------
Cost per ton
Vehicle Fuel savings CO2e reduced Cost per ton of the vehicle
Year program cost \b\ (million of the vehicle program with
\a\ ($millions) metric tons) program only fuel savings
($millions) \a\ \b\
----------------------------------------------------------------------------------------------------------------
2020............................ $15,600 -$35,700 160 $100 -$130
2030............................ 15,800 -79,800 310 50 -210
2040............................ 17,400 -119,300 400 40 -250
2050............................ 19,000 -171,200 510 40 -300
----------------------------------------------------------------------------------------------------------------
\a\ Costs here include vehicle compliance costs and do not include any fuel savings.
\b\ Fuel savings calculated using pre-tax fuel prices.

[[Page 25516]]

4. Reduction in Fuel Consumption and Its Impacts
a. What are the projected changes in fuel consumption?
The new CO2 standards will result in significant
improvements in the fuel efficiency of affected vehicles. Drivers of
those vehicles will see corresponding savings associated with reduced
fuel expenditures. EPA has estimated the impacts on fuel consumption
for both the tailpipe CO2 standards and the A/C credit
program. To do this, fuel consumption is calculated using both current
CO2 emission levels and the new CO2 standards.
The difference between these estimates represents the net savings from
the CO2 standards. Note that the total number of miles that
vehicles are driven each year is different under each of the control
case scenarios than in the reference case due to the ``rebound
effect,'' which is discussed in Section III.H.4.c. EPA also notes that
consumers who drive more than our average estimates for vehicle miles
traveled (VMT) will experience more fuel savings; consumers who drive
less than our average VMT estimates will experience less fuel savings.
The expected impacts on fuel consumption are shown in Table
III.H.4-1. The gallons shown in the tables reflect impacts from the new
CO2 standards, including the A/C credit program, and include
increased consumption resulting from the rebound effect.

Table III.H.4-1--Fuel Consumption Impacts of the Vehicle Standards and A/
C Credit Programs
[Million gallons]
------------------------------------------------------------------------
Year Total
------------------------------------------------------------------------
2012....................................................... 550
2013....................................................... 1,320
2014....................................................... 2,330
2015....................................................... 3,750
2016....................................................... 5,670
2020....................................................... 12,590
2030....................................................... 24,730
2040....................................................... 32,620
2050....................................................... 41,520
------------------------------------------------------------------------

b. What are the monetized fuel savings?
Using the fuel consumption estimates presented in Section
III.H.4.a, EPA can calculate the monetized fuel savings associated with
the CO2 standards. To do this, we multiply reduced fuel
consumption in each year by the corresponding estimated average fuel
price in that year, using the reference case taken from the AEO 2010
Early Release.\457\ AEO is the government consensus estimate used by
NHTSA and many other government agencies to estimate the projected
price of fuel. EPA has done this calculation using both the pre-tax and
post-tax fuel prices. Since the post-tax fuel prices are what consumers
pay, the fuel savings calculated using these prices represent the
savings consumers will see. The pre-tax fuel savings are those savings
that society will see. These results are shown in Table III.H.4-2. Note
that in Section III.H.10, EPA presents the benefit-cost of the rule
and, for that reason, presents only the pre-tax fuel savings.
---------------------------------------------------------------------------

\457\ Energy Information Administration. Annual Energy Outlook
2010 Early Release. Supplemental Transportation Tables. December
2009. http://www.eia.doe.gov/oiaf/aeo/supplement/sup_tran.xls.

Table III.H.4-2--Estimated Monetized Fuel Savings
[Millions of 2007 dollars]
------------------------------------------------------------------------
Fuel savings Fuel savings
Calendar year (pre-tax) (post-tax)
------------------------------------------------------------------------
2012.................................... $1,137 $1,400
2013.................................... 2,923 3,800
2014.................................... 5,708 6,900
2015.................................... 9,612 11,300
2016.................................... 14,816 17,400
2020.................................... 35,739 41,100
2030.................................... 79,838 89,100
2040.................................... 119,324 131,700
2050.................................... 171,248 186,300
NPV, 3%................................. 1,545,638 1,723,900
NPV, 7%................................. 672,629 755,700
------------------------------------------------------------------------

As shown in Table III.H.4-2, EPA is projecting that consumers would
realize very large fuel savings as a result of the standards contained
in this rule. As discussed further in Section III.H.1, it is a
conundrum from an economic perspective that these large fuel savings
have not been provided by automakers and purchased by consumers. A
number of behavioral and market phenomena may lead to this disparity
between the fuel economy that makes financial sense to consumers and
the fuel economy they purchase. Regardless how consumers make their
decisions on how much fuel economy to purchase, EPA expects that, in
the aggregate, they will gain these fuel savings, which will provide
actual money in consumers' pockets. We received considerable comment on
this issue, as discussed in Section III.H.1, and the issue is discussed
further in Chapter 8 of the RIA.
c. VMT Rebound Effect
The fuel economy rebound effect refers to the fraction of fuel
savings expected to result from an increase in vehicle fuel economy,
particularly one required by higher fuel efficiency standards, that is
offset by additional vehicle use. The increase in vehicle use occurs
because higher fuel economy reduces the fuel cost of driving, which is
typically the largest single component of the monetary cost of
operating a

[[Page 25517]]

vehicle, and vehicle owners respond to this reduction in operating
costs by driving slightly more.
For this rule, EPA is using an estimate of 10% for the rebound
effect. This value is based on the most recent time period analyzed in
the Small and Van Dender 2007 paper,\458\ and falls within the range of
the larger body of historical work on the rebound effect.\459\ Recent
work by David Greene on the rebound effect for light-duty vehicles in
the U.S. further supports the hypothesis that the rebound effect is
decreasing over time.\460\ If we were to use a dynamic estimate of the
future rebound effect, our analysis shows that the rebound effect could
be in the range of 5% or lower.\461\ The rebound effect is also further
discussed in Chapter 4 of the Joint TSD which reviews the relevant
literature and discusses in more depth the reasoning for the rebound
values used here.
---------------------------------------------------------------------------

\458\ Small, K. and K. Van Dender, 2007a. ``Fuel Efficiency and
Motor Vehicle Travel: The Declining Rebound Effect'', The Energy
Journal, vol. 28, no. 1, pp. 25-51 (Docket EPA-HQ-OAR-2009-0472-
0018).
\459\ Sorrell, S. and J. Dimitropoulos, 2007. ``UKERC Review of
Evidence for the Rebound Effect, Technical Report 2: Econometric
Studies'', UKERC/WP/TPA/2007/010, UK Energy Research Centre, London,
October (Docket EPA-HQ-OAR-2009-0472-0012).
\460\ Report by Kenneth A. Small of University of California at
Irvine to EPA, ``The Rebound Effect from Fuel Efficiency Standards:
Measurement and Projection to 2030'', June 12, 2009 (Docket EPA-HQ-
OAR-2009-0472-0002).
\461\ Revised Report by David Greene of Oak Ridge National
Laboratory to EPA, ``Rebound 2007: Analysis of National Light-Duty
Vehicle Travel Statistics,'' February 9, 2010 (Docket EPA-HQ-OAR-
2009-0472-0220). This paper has been accepted for an upcoming
special issue of Energy Policy, although the publication date has
not yet been determined.
---------------------------------------------------------------------------

We received several comments on the proposed value of the rebound
effect. The California Air Resources Board (CARB) and the New Jersey
Department of Environmental Protection supported the use of a 10%
rebound effect, although CARB encouraged EPA to consider lowering the
value to 5%. Other commenters, such as the Missouri Department of
Natural Resources, the International Council on Clean Transportation
(ICCT), the Center for Biological Diversity, and the Consumer
Federation of America, recommended using a lower rebound effect. ICCT
specifically recommended that the dynamic rebound effect methodology
utilized by Small & Van Dender was the most appropriate methodology,
which would support a rebound effect of 5% or lower. In contrast, the
National Association of Dealerships asserted that the rebound effect
should be higher (e.g., in the lower range of the 15-30% historical
range), but did not submit any data to support this claim.
While we appreciate the input provided by commenters, we did not
receive any new data or analysis to justify revising our initial
estimates of the rebound effect at this time. Based on the positive
comments we received, we will continue using the dynamic rebound effect
to help inform our estimate of the rebound effect in future
rulemakings. However, given the relatively new nature of this
analytical approach, we believe the larger body of historical studies
should also be considered when determining the value of the rebound
effect. As we described in the Technical Support Document, the more
recent literature suggests that the rebound effect is 10% or lower,
whereas the larger body of historical studies suggests a higher rebound
effect. Therefore, we will continue to use the 10% rebound effect for
this rulemaking. However, we plan to update our estimate of the rebound
effect in future rulemakings as new data becomes available.
We also invited comments on whether we should also explore other
alternatives for estimating the rebound effect, such as whether it
would be appropriate to use the price elasticity of demand for gasoline
to guide the choice of a value for the rebound effect. We received only
one comment on this issue from ICCT. In their comments, ICCT stated
that the short run elasticity can provide a useful point of comparison
for rebound effect estimates, but it should not be used to guide the
choice of a value for the rebound effect. Therefore, we have not
incorporated this metric into our analysis.
5. Impacts on U.S. Vehicle Sales and Payback Period
a. Vehicle Sales Impacts
This analysis compares two effects. On the one hand, the vehicles
will become more expensive, which would, by itself, discourage sales.
On the other hand, the vehicles will have improved fuel economy and
thus lower operating costs. If consumers do not accurately compare the
value of fuel savings with the increased cost of fuel economy
technology in their vehicle purchase decisions, as discussed in
Preamble III.H.1, they will continue to behave in this way after this
rule. If auto makers have accurately gauged how consumers consider fuel
economy when purchasing vehicles and have provided the amount that
consumers want in vehicles, then consumers should not be expected to
want the more fuel-efficient vehicles. After all, auto makers would
have provided as much fuel economy as consumers want. If, on the other
hand, auto makers underestimated consumer demand for fuel economy, as
suggested by some commenters and discussed in Preamble Section III.H.1
and RIA Section 8.1.2, then this rule may lead to production of more
desirable vehicles, and vehicle sales may increase. This assumption
implies that auto makers have missed some profit-making opportunities.
The methodology EPA used for estimating the impact on vehicle sales
is relatively straightforward, but makes a number of simplifying
assumptions. According to the literature, the price elasticity of
demand for vehicles is commonly estimated to be -1.0.\462\ In other
words, a one percent increase in the price of a vehicle would be
expected to decrease sales by one percent, holding all other factors
constant. For our estimates, EPA calculated the effect of an increase
in vehicle costs due to the GHG standards and assumes that consumers
will face the full increase in costs, not an actual (estimated) change
in vehicle price. (The estimated increases in vehicle cost due to the
rule are discussed in Section III.H.2.) This is a conservative
methodology, since an increase in cost may not pass fully into an
increase in market price in an oligopolistic industry such as the
automotive sector.\463\ EPA also notes that we have not used these
estimated sales impacts in the OMEGA Model.
---------------------------------------------------------------------------

\462\ Kleit A.N., 1990. ``The Effect of Annual Changes in
Automobile Fuel Economy Standards.'' Journal of Regulatory Economics
2: 151-172 (Docket EPA-HQ-OAR-2009-0472-0015); McCarthy, Patrick S.,
1996. ``Market Price and Income Elasticities of New Vehicle
Demands.'' Review of Economics and Statistics 78: 543-547 (Docket
EPA-HQ-OAR-2009-0472-0016); Goldberg, Pinelopi K., 1998. ``The
Effects of the Corporate Average Fuel Efficiency Standards in the
U.S.,'' Journal of Industrial Economics 46(1): 1-33 (Docket EPA-HQ-
OAR-2009-0472-0017).
\463\ See, for instance, Gron, Ann, and Deborah Swenson, 2000.
``Cost Pass-Through in the U.S. Automobile Market,'' Review of
Economics and Statistics 82: 316-324 (Docket EPA-HQ-OAR-2009-0472-
0007).
---------------------------------------------------------------------------

Although EPA uses the one percent price elasticity of demand for
vehicles as the basis for our vehicle sales impact estimates, we
assumed that the consumer would take into account both the higher
vehicle purchasing costs as well as some of the fuel savings benefits
when deciding whether to purchase a new vehicle. Therefore, the
incremental cost increase of a new vehicle would be offset by reduced
fuel expenditures over a certain period of time (i.e., the ``payback
period''). For the purposes of this rulemaking, EPA used a five-year
payback period, which is consistent with the length of a typical new
light-

[[Page 25518]]

duty vehicle loan.\464\ The one commenter on this analysis stated that
use of the five-year payback period was reasonable. This approach may
not accurately reflect the role of fuel savings in consumers' purchase
decisions, as the discussion in Section III.H.1 suggests. If consumers
consider fuel savings in a different fashion than modeled here, then
this approach will not accurately reflect the impact of this rule on
vehicle sales.
---------------------------------------------------------------------------

\464\ As discussed further in Section III.H.1, there is not a
consensus in the literature on how consumers consider fuel economy
in their vehicle purchases. Results are inconsistent, possibly due
to fuel economy not being a major focus of many of the studies, and
possibly due to sensitivity of results to modeling and data used. A
survey by Greene (Greene, David L. ``How Consumers Value Fuel
Economy: A Literature Review.'' EPA Report EPA-420-R-10-008, March
2010 (Docket EPA-HQ-OAR-2009-0472-11575)) finds that estimates in
the literature of the value that consumers place on fuel economy
when buying a vehicle range from negative--consumers would pay to
reduce fuel economy--to more than 1000 times the value of fuel
savings.
---------------------------------------------------------------------------

This increase in costs has other effects on consumers as well: if
vehicle prices increase, consumers will face higher insurance costs and
sales tax, and additional finance costs if the vehicle is bought on
credit. In addition, the resale value of the vehicles will increase.
EPA received no comments on these adjustments. The only change to these
adjustments between the NPRM and this discussion is an updating of the
interest rate on auto loans. EPA estimates that, with corrections for
these factors, the effect on consumer expenditures of the cost of the
new technology should be 0.914 times the cost of the technology at a 3%
discount rate, and 0.876 times the cost of the technology at a 7%
discount rate. The details of this calculation are in the RIA, Chapter
8.1.
Once the cost estimates are adjusted for these additional factors,
the fuel cost savings associated with the rule, discussed in Section
III.H.4, are subtracted to get the net effect on consumer expenditures
for a new vehicle. With the assumed elasticity of demand of -1, the
percent change in this ``effective price,'' estimated as the adjusted
increase in cost, is equal to the negative of the percent change in
vehicle purchases. The net effect of this calculation is in Table
III.H.5-1 and Table III.H.5-2. The values have changed slightly from
the NPRM, due to changes in fuel prices and fuel savings, technology
costs, and baseline vehicle sales projections, in addition to the
adjustment in financing costs.
The estimates provided in Table III.H.5-1 and Table III.H.5-2 are
meant to be illustrative rather than a definitive prediction. When
viewed at the industry-wide level, they give a general indication of
the potential impact on vehicle sales. As shown below, the overall
impact is positive and growing over time for both cars and trucks.
Because the fuel savings associated with this rule are expected to
exceed the technology costs, the effective prices of vehicles (the
adjusted increase in technology cost less the fuel savings over five
years) to consumers will fall, and consumers will buy more new
vehicles. As a result, the lower net cost of the vehicles is projected
to lead to an increase in sales for both cars and trucks.
As discussed above, this result depends on the assumption that more
fuel efficient vehicles that yield net consumer benefits over five
years would not otherwise be offered on the vehicle market due to
market failures on the part of vehicle manufacturers. If vehicles that
achieve the fuel economy standards prescribed by today's rulemaking
would already be available, but consumers chose not to purchase them,
then this rulemaking would not result in an increase in vehicle sales,
because it does not alter how consumers make decisions about which
vehicles to purchase. In addition, this analysis has not accounted for
a number of factors that might affect consumer vehicle purchases, such
as changing market conditions, changes in vehicle characteristics that
might accompany improvements in fuel economy, or consumers considering
a different ``payback period'' for their fuel economy purchases. If
consumers use a shorter payback period, the sales impacts will be less
positive, possibly negative; if consumers use a higher payback period,
the impacts will be more positive. Also, this is an aggregate analysis;
some individual consumers (those who drive less than estimated here)
will face lower net benefits, while others (who drive more than
estimated here) will have even greater savings. These complications add
considerable uncertainty to our vehicle sales impact analysis.

Table III.H.5-1--Vehicle Sales Impacts Using a 3% Discount Rate
----------------------------------------------------------------------------------------------------------------
Change in Change in
car sales % Change truck sales % Change
----------------------------------------------------------------------------------------------------------------
2012........................................................ 67,500 0.7 62,100 1.1
2013........................................................ 76,000 0.8 190,200 3.2
2014........................................................ 114,000 1.1 254,900 4.3
2015........................................................ 222,200 2.1 352,800 6.1
2016........................................................ 360,500 3.3 488,000 8.6
----------------------------------------------------------------------------------------------------------------

Table III.H.5-1 shows the impacts on new vehicle sales using a 3%
discount rate. The fuel savings over five years are always higher than
the technology costs. Although both cars and trucks show very small
effects initially, over time vehicle sales become increasingly
positive, as increased fuel prices make improved fuel economy more
desirable. The increases in sales for trucks are larger than the
increases for trucks (except in 2012) in both absolute numbers and
percentage terms.

Table III.H.5-2--New Vehicle Sales Impacts Using a 7% Discount Rate
----------------------------------------------------------------------------------------------------------------
Change in Change in
car sales % Change truck sales % Change
----------------------------------------------------------------------------------------------------------------
2012...................................................... 62,800 0.7 58,300 1
2013...................................................... 70,500 0.7 92,300 1.5
2014...................................................... 106,100 1 127,700 2.1

[[Page 25519]]

2015...................................................... 208,400 2 194,200 3.3
2016...................................................... 339,400 3.1 280,000 4.9
----------------------------------------------------------------------------------------------------------------

Table III.H.5-2 shows the impacts on new vehicle sales using a 7%
interest rate. While a 7% interest rate shows slightly lower impacts
than using a 3% discount rate, the results are qualitatively similar to
those using a 3% discount rate. Sales increase for every year. For both
cars and trucks, sales become increasingly positive over time, as
higher fuel prices make improved fuel economy more valuable. The car
market grows more than the truck market in absolute numbers, but less
on a percentage basis.
The effect of this rule on the use and scrappage of older vehicles
will be related to its effects on new vehicle prices, the fuel
efficiency of new vehicle models, and the total sales of new vehicles.
If the value of fuel savings resulting from improved fuel efficiency to
the typical potential buyer of a new vehicle outweighs the average
increase in new models' prices, sales of new vehicles will rise, while
scrappage rates of used vehicles will increase slightly. This will
cause the ``turnover'' of the vehicle fleet (i.e., the retirement of
used vehicles and their replacement by new models) to accelerate
slightly, thus accentuating the anticipated effect of the rule on
fleet-wide fuel consumption and CO2 emissions. However, if
potential buyers value future fuel savings resulting from the increased
fuel efficiency of new models at less than the increase in their
average selling price, sales of new vehicles will decline, as will the
rate at which used vehicles are retired from service. This effect will
slow the replacement of used vehicles by new models, and thus partly
offset the anticipated effects of this rule on fuel use and emissions.
Because the agencies are uncertain about how the value of projected
fuel savings from this rule to potential buyers will compare to their
estimates of increases in new vehicle prices, we have not attempted to
estimate explicitly the effects of the rule on scrappage of older
vehicles and the turnover of the vehicle fleet.
A detailed discussion of the vehicle sales impacts methodology is
provided in the Chapter 8 of EPA's RIA.
b. Consumer Payback Period and Lifetime Savings on New Vehicle
Purchases
Another factor of interest is the payback period on the purchase of
a new vehicle that complies with the new standards. In other words, how
long would it take for the expected fuel savings to outweigh the
increased cost of a new vehicle? For example, a new 2016 MY vehicle is
estimated to cost $948 more (on average, and relative to the reference
case vehicle) due to the addition of new GHG reducing technology (see
Section III.D.6 for details on this cost estimate). This new technology
will result in lower fuel consumption and, therefore, savings in fuel
expenditures (see Section III.H.10) for details on fuel savings). But
how many months or years would pass before the fuel savings exceed the
upfront cost of $948?
Table III.H.5-3 provides the answer to this question for a vehicle
purchaser who pays for the new vehicle upfront in cash (we discuss
later in this section the payback period for consumers who finance the
new vehicle purchase with a loan). The table uses annual miles driven
(vehicle miles traveled, or VMT) and survival rates consistent with the
emission and benefits analyses presented in Chapter 4 of the Joint TSD.
The control case includes rebound VMT but the reference case does not,
consistent with other parts of the analysis. Also included are fuel
savings associated with A/C controls (in the control case only). Not
included here are the likely A/C-related maintenance savings as
discussed in Chapter 2 of EPA's RIA. Further, this analysis does not
include other societal impacts such as the value of increased driving,
or noise, congestion and accidents since the focus is meant to be on
those factors consumers think about most while in the showroom
considering a new car purchase. Car/truck fleet weighting is handled as
described in Chapter 1 of the Joint TSD. As can be seen in the table,
it will take under 3 years (2 years and 7 months at a 3% discount rate,
2 years and 9 months at a 7% discount rate) for the cumulative
discounted fuel savings to exceed the upfront increase in vehicle cost.
More detail on this analysis can be found in Chapter 8 of EPA's RIA.

Table III.H.5-3--Payback Period on a 2016 MY New Vehicle Purchase via Cash
[2007 dollars]
----------------------------------------------------------------------------------------------------------------
Cumulative Cumulative
Increased Annual fuel discounted discounted
Year of ownership vehicle cost savings \b\ fuel savings fuel savings
\a\ at 3% at 7%
----------------------------------------------------------------------------------------------------------------
1............................................... $1,018 $424 $418 $410
2............................................... .............. $420 $820 $790
3............................................... .............. $414 $1,204 $1,139
4............................................... .............. $402 $1,567 $1,457
----------------------------------------------------------------------------------------------------------------
\a\ Increased vehicle cost due to the rule is $948; the value here includes nationwide average sales tax of 5.3%
and increased insurance premiums of 1.98%; both of these percentages are discussed in Section 8.1.1 of EPA's
RIA.
\b\ Calculated using AEO 2010 Early Release reference case fuel price including taxes.

However, most people purchase a new vehicle using credit rather
than paying cash up front. The typical car loan today is a five year,
60 month loan. As of February 9, 2010, the national average interest
rate for a 5 year new car loan was 6.54 percent. If the increased
vehicle cost is spread out over 5 years at 6.54 percent, the analysis
would look like that shown in Table III.H.5-4. As can be seen in this
table, the fuel savings immediately outweigh the

[[Page 25520]]

increased payments on the car loan, amounting to $177 in discounted net
savings (3% discount rate) in the first year and similar savings for
the next two years before reduced VMT starts to cause the fuel savings
to fall. Results are similar using a 7% discount rate. This means that
for every month that the average owner is making a payment for the
financing of the average new vehicle their monthly fuel savings would
be greater than the increase in the loan payments. This amounts to a
savings on the order of $9 to $15 per month throughout the duration of
the 5 year loan. Note that in year six when the car loan is paid off,
the net savings equal the fuel savings (as would be the case for the
remaining years of ownership).

Table III.H.5-4--Payback Period on a 2016 MY New Vehicle Purchase via Credit
[2007 dollars]
----------------------------------------------------------------------------------------------------------------
Increased Annual Annual
Year of ownership vehicle cost Annual fuel discounted net discounted net
\a\ savings \b\ savings at 3% savings at 7%
----------------------------------------------------------------------------------------------------------------
1............................................... $245 $424 $177 $173
2............................................... $245 $420 $167 $158
3............................................... $245 $414 $157 $142
4............................................... $245 $402 $142 $124
5............................................... $245 $391 $127 $107
6............................................... $0 $374 $318 $258
----------------------------------------------------------------------------------------------------------------
\a\ This uses the same increased cost as Table III.H.4-3 but spreads it out over 5 years assuming a 5 year car
loan at 6.54 percent.
\b\ Calculated using AEO 2010 Early Release reference case fuel price including taxes.

The lifetime fuel savings and net savings can also be calculated
for those who purchase the vehicle using cash and for those who
purchase the vehicle with credit. This calculation applies to the
vehicle owner who retains the vehicle for its entire life and drives
the vehicle each year at the rate equal to the national projected
average. The results are shown in Table III.H.5-5. In either case, the
present value of the lifetime net savings is greater than $3,100 at a
3% discount rate, or $2,300 at a 7% discount rate.

Table III.H.5-5--Lifetime Discounted Net Savings on a 2016 MY New Vehicle Purchase
[2007 dollars]
----------------------------------------------------------------------------------------------------------------
Lifetime
Increased discounted Lifetime
Purchase option discounted fuel savings discounted net
vehicle cost \b\ savings
----------------------------------------------------------------------------------------------------------------
3% discount rate
----------------------------------------------------------------------------------------------------------------
Cash............................................................ $1,018 $4,306 $3,303
Credit \a\...................................................... 1,140 4,306 3,166
----------------------------------------------------------------------------------------------------------------
7% discount rate
----------------------------------------------------------------------------------------------------------------
Cash............................................................ 1,018 3,381 2,396
Credit \a\...................................................... 1,040 3,381 2,340
----------------------------------------------------------------------------------------------------------------
\a\ Assumes a 5 year loan at 6.54 percent.
\b\ Fuel savings here were calculated using AEO 2010 Early Release reference case fuel price including taxes.

Note that throughout this consumer payback discussion, the average
number of vehicle miles traveled per year has been used. Drivers who
drive more miles than the average would incur fuel related savings more
quickly and, therefore, the payback would come sooner. Drivers who
drive fewer miles than the average would incur fuel related savings
more slowly and, therefore, the payback would come later.
6. Benefits of Reducing GHG Emissions
a. Social Cost of Carbon
In today's final rule, EPA and NHTSA assigned a dollar value to
reductions in CO2 emissions using the marginal dollar value
of climate-related damages resulting from carbon emissions, also
referred to as ``social cost of carbon'' (SCC). The SCC estimates used
in today's rule were recently developed by an interagency process, in
which EPA and NHTSA participated. As part of the interagency group, EPA
and NHTSA have critically evaluated the new SCC estimates and endorse
them for use in these regulatory analyses, for the reasons presented
below. The SCC TSD, Social Cost of Carbon for Regulatory Impact
Analysis Under Executive Order 12866, presents a more detailed
description of the methodology used to generate the new estimates, the
underlying assumptions, and the limitations of the new SCC estimates.
Under Executive Order 12866, agencies are required, to the extent
permitted by law, ``to assess both the costs and the benefits of the
intended regulation and, recognizing that some costs and benefits are
difficult to quantify, propose or adopt a regulation only upon a
reasoned determination that the benefits of the intended regulation
justify its costs.'' The purpose of the SCC estimates presented here is
to incorporate the social benefits of reducing carbon dioxide
(CO2) emissions from light-duty vehicles into a cost-benefit
analysis of this final rule, which has a small, or ``marginal,'' impact
on cumulative global emissions. The estimates are presented with an
acknowledgement of the many

[[Page 25521]]

uncertainties involved and with a clear understanding that they should
be updated over time to reflect increasing knowledge of the science and
economics of climate impacts.
The interagency process that developed these SCC estimates involved
a group of technical experts from numerous agencies, which met on a
regular basis to consider public comments, explore the technical
literature in relevant fields, and discuss key model inputs and
assumptions. The main objective of this process was to develop a range
of SCC values using a defensible set of input assumptions grounded in
the existing scientific and economic literatures. In this way, key
uncertainties and model differences transparently and consistently
inform the range of SCC estimates used in this rulemaking process.
The interagency group selected four SCC values for use in
regulatory analyses, which EPA and NHTSA have applied to this final
rule. Three values are based on the average SCC from three integrated
assessment models, at discount rates of 2.5, 3, and 5 percent. The
fourth value, which represents the 95th percentile SCC estimate across
all three models at a 3 percent discount rate, is included to represent
higher-than-expected impacts from temperature change further out in the
tails of the SCC distribution.

Table III.H.6-1--Social Cost of CO2, 2010--2050a
[in 2007 dollars]
------------------------------------------------------------------------
Discount Rate
Year -------------------------------------------
5% Avg 3% Avg 2.5% Avg 3% 95th
------------------------------------------------------------------------
2010........................ 5 21 35 65
2015........................ 6 24 38 73
2020........................ 7 26 42 81
2025........................ 8 30 46 90
2030........................ 10 33 50 100
2035........................ 11 36 54 110
2040........................ 13 39 58 119
2045........................ 14 42 62 128
2050........................ 16 45 65 136
------------------------------------------------------------------------
\a\ The SCC estimates presented above have been rounded to nearest
dollar for consistency with the benefits analysis. The SCC TSD
presents estimates rounded to the nearest tenth of a cent.

i. Monetizing Carbon Dioxide Emissions
The ``social cost of carbon'' (SCC) is an estimate of the monetized
damages associated with an incremental increase in carbon emissions in
a given year. It is intended to include (but is not limited to) changes
in net agricultural productivity, human health, property damages from
increased flood risk, and the value of ecosystem services. We report
estimates of the social cost of carbon in dollars per metric ton of
carbon dioxide throughout this document.
When attempting to assess the incremental economic impacts of
carbon dioxide emissions, the analyst faces a number of serious
challenges. A 2009 report from the National Academies of Science points
out that any assessment will suffer from uncertainty, speculation, and
lack of information about (1) future emissions of greenhouse gases, (2)
the effects of past and future emissions on the climate system, (3) the
impact of changes in climate on the physical and biological
environment, and (4) the translation of these environmental impacts
into economic damages.\465\ As a result, any effort to quantify and
monetize the harms associated with climate change will raise serious
questions of science, economics, and ethics and should be viewed as
provisional.
---------------------------------------------------------------------------

\465\ National Research Council (2009). Hidden Costs of Energy:
Unpriced Consequences of Energy Production and Use. National
Academies Press.
---------------------------------------------------------------------------

Despite the serious limits of both quantification and monetization,
SCC estimates can be useful in estimating the social benefits of
reducing carbon dioxide emissions. Under Executive Order 12866,
agencies are required, to the extent permitted by law, ``to assess both
the costs and the benefits of the intended regulation and, recognizing
that some costs and benefits are difficult to quantify, propose or
adopt a regulation only upon a reasoned determination that the benefits
of the intended regulation justify its costs.'' EPA and NHTSA have used
the SCC estimates to incorporate social benefits from reducing carbon
dioxide emissions from light-duty vehicles into a cost-benefit analysis
of this final rule, which has a small, or ``marginal,'' impact on
cumulative global emissions. Most Federal regulatory actions can be
expected to have marginal impacts on global emissions.
For policies that have marginal impacts on global emissions, the
benefits from reduced (or costs from increased) emissions in any future
year can be estimated by multiplying the change in emissions in that
year by the SCC value appropriate for that year. The net present value
of the benefits can then be calculated by multiplying each of these
future benefits by an appropriate discount factor and summing across
all affected years. This approach assumes that the marginal damages
from increased emissions are constant for small departures from the
baseline emissions path, an approximation that is reasonable for
policies that have effects on emissions that are small relative to
cumulative global carbon dioxide emissions. For policies that have a
large (non-marginal) impact on global cumulative emissions, there is a
separate question of whether the SCC is an appropriate tool for
calculating the benefits of reduced emissions; we do not attempt to
answer that question here.
As noted above, the interagency group convened on a regular basis
to consider public comments, explore the technical literature in
relevant fields, and discuss key inputs and assumptions in order to
generate SCC estimates. In addition to EPA and NHTSA, agencies that
actively participated in the interagency process included the
Departments of Agriculture, Commerce, Energy, and Treasury. This
process was convened by the Council of Economic Advisers and the Office
of Management and Budget, with active participation and regular input
from the Council on Environmental Quality, National Economic Council,
Office of Energy and Climate Change, and Office of Science and
Technology Policy. The main objective of this process was to develop a
range of SCC values using a defensible

[[Page 25522]]

set of input assumptions that are grounded in the existing literature.
In this way, key uncertainties and model differences can more
transparently and consistently inform the range of SCC estimates used
in the rulemaking process.
The interagency group selected four global SCC estimates for use in
regulatory analyses. For 2010, these estimates are $5, $21, $35, and
$65 (in 2007 dollars). The first three estimates are based on the
average SCC across models and socio-economic and emissions scenarios at
the 5, 3, and 2.5 percent discount rates, respectively. The fourth
value is included to represent the higher-than-expected impacts from
temperature change further out in the tails of the SCC distribution.
For this purpose, we use the SCC value for the 95th percentile at a 3
percent discount rate. The central value is the average SCC across
models at the 3 percent discount rate. For purposes of capturing the
uncertainties involved in regulatory impact analysis, we emphasize the
importance and value of considering the full range. These SCC estimates
also grow over time. For instance, the central value increases to $24
per ton of CO2 in 2015 and $26 per ton of CO2 in
2020. See the SCC TSD for the full range of annual SCC estimates from
2010 to 2050.
These new SCC estimates represent global measures and the center of
our current attention because of the distinctive nature of the climate
change problem. The climate change problem is highly unusual in at
least two respects. First, it involves a global externality: Emissions
of most greenhouse gases contribute to damages around the world even
when they are emitted in the United States. Consequently, to address
the global nature of the problem, the SCC must incorporate the full
(global) damages caused by GHG emissions. Second, climate change
presents a problem that the United States alone cannot solve. Even if
the United States were to reduce its greenhouse gas emissions to zero,
that step would be far from enough to avoid substantial climate change.
Other countries would also need to take action to reduce emissions if
significant changes in the global climate are to be avoided.
It is important to emphasize that the interagency process is
committed to updating these estimates as the science and economic
understanding of climate change and its impacts on society improves
over time. Specifically, the interagency group has set a preliminary
goal of revisiting the SCC values within two years or at such time as
substantially updated models become available, and to continue to
support research in this area. In the meantime, the interagency group
will continue to explore the issues raised in the SCC TSD and consider
public comments as part of the ongoing interagency process.
ii. Social Cost of Carbon Values Used in Past Regulatory Analyses
To date, economic analyses for Federal regulations have used a wide
range of values to estimate the benefits associated with reducing
carbon dioxide emissions. In the final model year 2011 CAFE rule, the
Department of Transportation (DOT) used both a ``domestic'' SCC value
of $2 per ton of CO2 and a ``global'' SCC value of $33 per
ton of CO2 for 2007 emission reductions (in 2007 dollars),
increasing both values at 2.4 percent per year. It also included a
sensitivity analysis at $80 per ton of CO2. A domestic SCC
value is meant to reflect the value of damages in the United States
resulting from a unit change in carbon dioxide emissions, while a
global SCC value is meant to reflect the value of damages worldwide.
A 2008 regulation proposed by DOT assumed a domestic SCC value of
$7 per ton CO2 (in 2006 dollars) for 2011 emission
reductions (with a range of $0-$14 for sensitivity analysis), also
increasing at 2.4 percent per year. A regulation finalized by DOE in
October of 2008 used a domestic SCC range of $0 to $20 per ton
CO2 for 2007 emission reductions (in 2007 dollars). In
addition, EPA's 2008 Advance Notice of Proposed Rulemaking for
Greenhouse Gases identified what it described as ``very preliminary''
SCC estimates subject to revision. EPA's global mean values were $68
and $40 per ton CO2 for discount rates of approximately 2
percent and 3 percent, respectively (in 2006 dollars for 2007
emissions).
In 2009, an interagency process was initiated to offer a
preliminary assessment of how best to quantify the benefits from
reducing carbon dioxide emissions. To ensure consistency in how
benefits are evaluated across agencies, the Administration sought to
develop a transparent and defensible method, specifically designed for
the rulemaking process, to quantify avoided climate change damages from
reduced CO2 emissions. The interagency group did not
undertake any original analysis. Instead, it combined SCC estimates
from the existing literature to use as interim values until a more
comprehensive analysis could be conducted.
The outcome of the preliminary assessment by the interagency group
was a set of five interim values: Global SCC estimates for 2007 (in
2006 dollars) of $55, $33, $19, $10, and $5 per ton of CO2.
The $33 and $5 values represented model-weighted means of the published
estimates produced from the most recently available versions of three
integrated assessment models (DICE, PAGE, and FUND) at approximately 3
and 5 percent discount rates.\466\ The $55 and $10 values were derived
by adjusting the published estimates for uncertainty in the discount
rate (using factors developed by Newell and Pizer (2003)) at 3 and 5
percent discount rates, respectively.\467\ The $19 value was chosen as
a central value between the $5 and $33 per ton estimates. All of these
values were assumed to increase at 3 percent annually to represent
growth in incremental damages over time as the magnitude of climate
change increases.
---------------------------------------------------------------------------

\466\ The DICE (Dynamic Integrated Climate and Economy) model by
William Nordhaus evolved from a series of energy models and was
first presented in 1990 (Nordhaus and Boyer 2000, Nordhaus 2008).
The PAGE (Policy Analysis of the Greenhouse Effect) model was
developed by Chris Hope in 1991 for use by European decision-makers
in assessing the marginal impact of carbon emissions (Hope 2006,
Hope 2008). The FUND (Climate Framework for Uncertainty,
Negotiation, and Distribution) model, developed by Richard Tol in
the early 1990s, originally to study international capital transfers
in climate policy, is now widely used to study climate impacts
(e.g., Tol 2002a, Tol 2002b, Anthoff et al. 2009, Tol 2009).
\467\ Newell, R., and W. Pizer. 2003. Discounting the distant
future: How much do uncertain rates increase valuations? Journal of
Environmental Economics and Management 46: 52-71.
---------------------------------------------------------------------------

These interim values represent the first sustained interagency
effort within the U.S. Government to develop an SCC for use in
regulatory analysis. The results of this preliminary effort were
presented in several proposed and final rules and were offered for
public comment in connection with proposed rules. In particular, EPA
and NHTSA used the interim SCC estimates in the joint proposal leading
to this final rule.
iii. Approach and Key Assumptions
Since the release of the interim values, interagency group has
reconvened on a regular basis to generate improved SCC estimates, which
EPA and NHTSA used in this final rule. Specifically, the group has
considered public comments and further explored the technical
literature in relevant fields. The general approach to estimating SCC
values was to run the three integrated assessment models (FUND, DICE,
and PAGE) using the following inputs agreed upon by the interagency
group:
A Roe and Baker distribution for the climate sensitivity
parameter bounded between 0 and 10 with a median of 3 [deg]C and a
cumulative probability between 2 and 4.5 [deg]C of two-thirds.\468\
---------------------------------------------------------------------------

\468\ Roe, G., and M. Baker. 2007. ``Why is climate sensitivity
so unpredictable?'' Science 318:629-632.

---------------------------------------------------------------------------

[[Page 25523]]

Five sets of GDP, population and carbon emissions
trajectories based on the recent Stanford Energy Modeling Forum, EMF-
22.
Constant annual discount rates of 2.5, 3, and 5 percent.

The SCC TSD presents a summary of the results and details, the modeling
exercise and the choices and assumptions that underlie the resulting
estimates of the SCC. The complete model results are available in the
docket for this final rule [EPA-HQ-OAR-2009-0472].
It is important to recognize that a number of key uncertainties
remain, and that current SCC estimates should be treated as provisional
and revisable since they will evolve with improved scientific and
economic understanding. The interagency group also recognizes that the
existing models are imperfect and incomplete. The National Academy of
Science (2009) points out that there is tension between the goal of
producing quantified estimates of the economic damages from an
incremental ton of carbon and the limits of existing efforts to model
these effects. The SCC TSD highlights a number of concerns and problems
that should be addressed by the research community, including research
programs housed in many of the agencies participating in the
interagency process to estimate the SCC.
The U.S. Government will periodically review and reconsider
estimates of the SCC used for cost-benefit analyses to reflect
increasing knowledge of the science and economics of climate impacts,
as well as improvements in modeling. In this context, statements
recognizing the limitations of the analysis and calling for further
research take on exceptional significance. The interagency group offers
the new SCC values with all due humility about the uncertainties
embedded in them and with a sincere promise to continue work to improve
them.
iv. Use of New SCC Estimates To Calculate GHG Benefits for This Final
Rule
The table below summarizes the total GHG benefits for the lifetime
of the rule, which are calculated by using the four new SCC values.
Specifically, EPA calculated the total monetized benefits in each year
by multiplying the marginal benefits estimates per metric ton of
CO2 (the SCC) by the reductions in CO2 for that
year.

Table III.H.6-2--Monetized CO2 Benefits of Vehicle Program, CO2 Emissions a b
[Million 2007$]
----------------------------------------------------------------------------------------------------------------
Benefits
CO2 emissions ---------------------------------------------------------------
reduction 95th
Year (Million Avg SCC at 5% Avg SCC at 3% Avg SCC at percentile SCC
metric tons) ($5-$16) c ($21-$45) c 2.5% ($35-$65) at 3% ($65-
c $136) c
----------------------------------------------------------------------------------------------------------------
2020............................ 139 $900 $3,700 $5,800 $11,000
2030............................ 273 2,700 8,900 14,000 27,000
2040............................ 360 4,600 14,000 21,000 43,000
2050............................ 459 7,200 21,000 30,000 62,000
----------------------------------------------------------------------------------------------------------------
\a\ Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
\b\ Numbers may not compute exactly from Tables III.H.6-1 and III.H.6-2 due to rounding.
\c\ As noted above, SCC increases over time; tables lists ranges for years 2010 through 2050. See Table III.H.6-
1 for the SCC estimates corresponding to the years in this table.

b. Summary of the Response to Comments
EPA and NHTSA received extensive public comments about the
scientific, economic, and ethical issues involved in estimating the
SCC, including the proposed rule's estimates of the value of emissions
reductions from new cars and trucks.\469\ In particular, the comments
addressed the methodology used to derive the interim SCC estimates,
limitations of integrated assessment models, discount rate selection,
treatment of uncertainty and catastrophic impacts, use of global and
domestic SCC, and the presentation and use of SCC estimates. The rest
of this preamble section briefly summarizes EPA's response to the
comments; the Response to Comments document provides the complete
responses to all comments received.
---------------------------------------------------------------------------

\469\ EPA estimated GHG benefits in the proposed rule using a
set of interim SCC values developed by an interagency group, in
which EPA and NHTSA participated. As discussed in the SCC TSD, the
interagency group selected the interim estimates from the existing
literature and agreed to use those interim estimates in regulatory
analyses until it could develop a more comprehensive
characterization of the SCC.
---------------------------------------------------------------------------

EPA received extensive comments about the methodology and discount
rates used to derive the interim SCC estimates. While one commenter
from the auto industry noted that the interim methodology was
acceptable given available data, many commenters (representing academic
and environmental organizations) expressed concerns that the filters
were too narrow, stated that model-weighting averaging was
inappropriate, and recommended that EPA use lower discount rates. These
commenters also discussed alternative approaches to select discount
rates and generally recommended that EPA use lower rates to give more
weight to climate damages experienced by future generations.
For the final rule, EPA conducted new analyses of SCC. EPA did not
continue with its interim approach to derive estimates from the
existing literature and instead conducted new model runs that produced
a vast amount of SCC data at three separate certainty-equivalent
discount rates (2.5, 3, and 5 percent). As discussed further in the SCC
TSD, this modeling exercise resulted in a fuller distribution of SCC
estimates and better accounted for uncertainty through a Monte Carlo
analysis. Comments on specific issues are addressed in the Response to
Comments document.
EPA received comments on the limitations of the integrated
assessment models concluding that the selection of models and reliance
on the model authors' datasets contributed to the downward bias of the
interim SCC estimates. In this final rule, EPA relied on the default
values in each model for the remaining parameter; research gaps

[[Page 25524]]

and practical constraints required EPA to limit its modification of the
models to socioeconomic and emissions scenarios, climate sensitivity,
and discount rate. While EPA recognizes that the models' translations
of physical impacts to economic values are incomplete, approximate, and
highly uncertain, it regards them as the best currently available
representations. EPA also considered, for each model, the treatment of
uncertainty, catastrophic impacts, and omitted impacts, and as
discussed in the SCC TSD and the Response to Comments document, used
best available information and techniques to quantify such impacts as
feasible and supplemented the SCC with qualitative assessments.
Comments on specific issues are addressed in the Response to Comments
document.
Six commenters, representing academia and environmental
organizations, supported the proposed rule's preference for global SCC
estimates while several industry groups stated that under the Clean Air
Act, EPA is prohibited from using global estimates. EPA agrees that a
global measure of GHG mitigation benefits is both appropriate and
lawful for EPA to consider in evaluating the benefits of GHG emissions
standards adopted under section 202(a). Global climate change
represents a problem that the United States cannot solve alone without
global action, and for a variety of reasons there is a value to the
U.S. from domestic emissions reductions that reduce the harm occurring
globally. This is not exercise of regulatory authority over conduct
occurring overseas, but instead is a reasonable exercise of discretion
in how to place a monetary value on a reduction in domestic emissions.
See the Response to Comments document for a complete discussion of this
issue.
Finally, EPA received various comments regarding the presentation
of the SCC methodology and resulting estimates. EPA has responded to
these concerns by presenting a detailed discussion about the
methodology, including key model assumptions, as well as uncertainties
and research gaps associated with the SCC estimates and the
implications for the SCC estimates. Among these key assumptions and
uncertainties are issues involving discount rates, climate sensitivity
and socioeconomic scenario assumptions, incomplete treatment of
potential catastrophic impacts, incomplete treatment of non-
catastrophic impacts, uncertainty in extrapolation of damages to high
temperatures, incomplete treatment of adaptation and technological
change, and assumptions about risk aversion to high-impact outcomes
(see SCC TSD).
7. Non-Greenhouse Gas Health and Environmental Impacts
This section presents EPA's analysis of the non-GHG health and
environmental impacts that can be expected to occur as a result of the
light-duty vehicle GHG rule. GHG emissions are predominantly the
byproduct of fossil fuel combustion processes that also produce
criteria and hazardous air pollutants. The vehicles that are subject to
the standards are also significant sources of mobile source air
pollution such as direct PM, NOX, VOCs and air toxics. The
standards will affect exhaust emissions of these pollutants from
vehicles. They will also affect emissions from upstream sources related
to changes in fuel consumption. Changes in ambient ozone,
PM2.5, and air toxics that will result from the standards
are expected to affect human health in the form of premature deaths and
other serious human health effects, as well as other important public
health and welfare effects.
As many commenters noted, it is important to quantify the health
and environmental impacts associated with the final rule because a
failure to adequately consider these ancillary co-pollutant impacts
could lead to an incorrect assessment of their net costs and benefits.
Moreover, co-pollutant impacts tend to accrue in the near term, while
any effects from reduced climate change mostly accrue over a timeframe
of several decades or longer.
This section is split into two sub-sections: The first presents the
PM- and ozone-related health and environmental impacts associated with
the final rule in calendar year (CY) 2030; the second presents the PM-
related benefits-per-ton values used to monetize the PM-related co-
benefits associated with the model year (MY) analysis of the final
rule.\470\
---------------------------------------------------------------------------

\470\ EPA typically analyzes rule impacts (emissions, air
quality, costs and benefits) in the year in which they occur; for
this analysis, we selected 2030 as a representative future year. We
refer to this analysis as the ``Calendar Year'' (CY) analysis. EPA
also conducted a separate analysis of the impacts over the model
year lifetimes of the 2012 through 2016 model year vehicles. We
refer to this analysis as the ``Model Year'' (MY) analysis. In
contrast to the CY analysis, the MY lifetime analysis shows the
lifetime impacts of the program on each of these MY fleets over the
course of its lifetime.
---------------------------------------------------------------------------

a. Quantified and Monetized Non-GHG Human Health Benefits of the 2030
Calendar Year (CY) Analysis
This analysis reflects the impact of the final light-duty GHG rule
in 2030 compared to a future-year reference scenario without the rule
in place. Overall, we estimate that the final rule will lead to a net
decrease in PM2.5-related health impacts (see Section
III.G.5 of this preamble for more information about the air quality
modeling results). While the PM-related air quality impacts are
relatively small, the decrease in population-weighted national average
PM2.5 exposure results in a net decrease in adverse PM-
related human health impacts (the decrease in national population-
weighted annual average PM2.5 is 0.0036 [mu]g/
m³).
The air quality modeling (discussed in Section III.G.5) projects
very small increases in ozone concentrations in many areas, but these
are driven by the ethanol production volumes mandated by the recently
finalized RFS2 rule and are not due to the standards finalized in this
rule. While the ozone-related impacts are very small, the increase in
population-weighted national average ozone exposure results in a small
increase in ozone-related health impacts (population-weighted maximum
8-hour average ozone increases by 0.0104 ppb).
We base our analysis of the final rule's impact on human health in
2030 on peer-reviewed studies of air quality and human health
effects.^{471 472} These methods are described in more detail
in the RIA that accompanies this action. Our benefits methods are also
consistent with recent rulemaking analyses such as the proposed
Portland Cement National Emissions Standards for Hazardous Air
Pollutants (NESHAP) RIA,\473\ the final NO2 NAAQS,\474\ and
the final Category 3 Marine Engine rule.\475\ To model the

[[Page 25525]]

ozone and PM air quality impacts of the final rule, we used the
Community Multiscale Air Quality (CMAQ) model (see Section III.G.5).
The modeled ambient air quality data serves as an input to the
Environmental Benefits Mapping and Analysis Program (BenMAP).\476\
BenMAP is a computer program developed by the U.S. EPA that integrates
a number of the modeling elements used in previous analyses (e.g.,
interpolation functions, population projections, health impact
functions, valuation functions, analysis and pooling methods) to
translate modeled air concentration estimates into health effects
incidence estimates and monetized benefits estimates.
---------------------------------------------------------------------------

\471\ U.S. Environmental Protection Agency. (2006). Final
Regulatory Impact Analysis (RIA) for the National Ambient Air
Quality Standards for Particulate Matter. Prepared by: Office of Air
and Radiation. Retrieved March 26, 2009 at http://www.epa.gov/ttn/
ecas/ria.html. EPA-HQ-OAR-2009-0472-0240.
\472\ U.S. Environmental Protection Agency. (2008). Final Ozone
NAAQS Regulatory Impact Analysis. Prepared by: Office of Air and
Radiation, Office of Air Quality Planning and Standards. Retrieved
March 26, 2009 at http://www.epa.gov/ttn/ecas/ria.html. EPA-HQ-OAR-
2009-0472-0238.
\473\ U.S. Environmental Protection Agency (U.S. EPA). 2009.
Regulatory Impact Analysis: National Emission Standards for
Hazardous Air Pollutants from the Portland Cement Manufacturing
Industry. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. April. Available on the Internet at http://
www.epa.gov/ttn/ecas/regdata/RIAs/portlandcementria_4-20-09.pdf.
Accessed March 15, 2010. EPA-HQ-OAR-2009-0472-0241.
\474\ U.S. Environmental Protection Agency (U.S. EPA). 2010.
Final NO2 NAAQS Regulatory Impact Analysis (RIA). Office
of Air Quality Planning and Standards, Research Triangle Park, NC.
April. Available on the Internet at http://www.epa.gov/ttn/ecas/
regdata/RIAs/FinalNO2RIAfulldocument.pdf. Accessed March 15, 2010.
EPA-HQ-OAR-2009-0472-0237.
\475\ U.S. Environmental Protection Agency. 2009. Regulatory
Impact Analysis: Control of Emissions of Air Pollution from Category
3 Marine Diesel Engines. EPA-420-R-09-019, December 2009. Prepared
by Office of Air and Radiation. http://www.epa.gov/otaq/regs/
nonroad/marine/ci/420r09019.pdf. Accessed February 9, 2010. EPA-HQ-
OAR-2009-0472-0283.
\476\ Information on BenMAP, including downloads of the
software, can be found at http://www.epa.gov/ttn/ecas/
benmodels.html.
---------------------------------------------------------------------------

The range of total monetized ozone- and PM-related health impacts
is presented in Table III.H.7-1. We present total benefits based on the
PM- and ozone-related premature mortality function used. The benefits
ranges therefore reflect the addition of each estimate of ozone-related
premature mortality (each with its own row in Table III.H.7-1) to
estimates of PM-related premature mortality. These estimates represent
EPA's preferred approach to characterizing a best estimate of benefits.
As is the nature of Regulatory Impact Analyses (RIAs), the assumptions
and methods used to estimate air quality benefits evolve to reflect the
Agency's most current interpretation of the scientific and economic
literature.

Table III.H.7-1--Estimated 2030 Monetized PM- and Ozone-Related Health Benefits a
----------------------------------------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------
2030 Total Ozone and PM Benefits--PM Mortality Derived from American Cancer Society Analysis and Six-Cities
Analysis \a\
----------------------------------------------------------------------------------------------------------------
Premature Ozone Mortality Function Reference Total Benefits Total Benefits
(Millions, 2007$, 3% (Millions, 2007$, 7%
Discount Rate) b c d Discount Rate) b c d
----------------------------------------------------------------------------------------------------------------
Multi-city analyses.................. Bell et al., 2004...... Total: $510-$1,300..... Total: $460-$1,200
PM: $550-$1,300........ PM: $500-$1,200
Ozone: -$40............ Ozone: -$40
Huang et al., 2005..... Total: $490-$1,300..... Total: $440-$1,200
PM: $550-$1,300........ PM: $500-$1,200
Ozone: -$64............ Ozone: -$64
Schwartz, 2005......... Total: $490-$1,300..... Total: $440-$1,200
PM: $550-$1,300........ PM: $500-$1,200
Ozone: -$60............ Ozone: -$60
Meta-analyses........................ Bell et al., 2005...... Total: $430-$1,200..... Total: $380-$1,100
PM: $550-$1,300........ PM: $500-$1,200
Ozone: -$120........... Ozone: -$120
Ito et al., 2005....... Total: $380-$1,200..... Total: $330-$1,000
PM: $550-$1,300........ PM: $500-$1,200
Ozone: -$170........... Ozone: -$170
Levy et al., 2005...... Total: $380-$1,200..... Total: $330-$1,000
PM: $550-$1,300........ PM: $500-$1,200
Ozone: -$170........... Ozone: -$170
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ Total includes premature mortality-related and morbidity-related ozone and PM2.5 benefits. Range was
developed by adding the estimate from the ozone premature mortality function to the estimate of PM2.5-related
premature mortality derived from either the ACS study (Pope et al., 2002) \477\ or the Six-Cities study (Laden
et al., 2006).\478\
\b\ Note that total benefits presented here do not include a number of unquantified benefits categories. A
detailed listing of unquantified health and welfare effects is provided in Table III.H.7-2.
\c\ Results reflect the use of both a 3 and 7 percent discount rate, as recommended by EPA's Guidelines for
Preparing Economic Analyses and OMB Circular A-4. Results are rounded to two significant digits for ease of
presentation and computation.
\d\ Negatives indicate a disbenefit, or an increase in health effect incidence.

The benefits in Table III.H.7-1 include all of the human health
impacts we are able to quantify and monetize at this time. However, the
full complement of human health and welfare effects associated with PM
and ozone remain unquantified because of current limitations in methods
or available data. We have not quantified a number of known or
suspected health effects linked with ozone and PM for which appropriate
health impact functions are not available or which do not provide
easily interpretable outcomes (e.g., changes in heart rate
variability). Additionally, we are unable to quantify a number of known
welfare effects, including reduced acid and particulate deposition
damage to cultural monuments and other materials, and environmental
benefits due to reductions of impacts of eutrophication in coastal
areas. These are listed in Table III.H.7-2. As a result, the health
benefits quantified in this section are likely underestimates of the
total benefits attributable to the final rule.
---------------------------------------------------------------------------

\477\ Pope, C.A., III, R.T. Burnett, M.J. Thun, E.E. Calle, D.
Krewski, K. Ito, and G.D. Thurston (2002). ``Lung Cancer,
Cardiopulmonary Mortality, and Long-term Exposure to Fine
Particulate Air Pollution.'' Journal of the American Medical
Association 287:1132-1141. EPA-HQ-OAR-2009-0472-0263.
\478\ Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery
(2006). Reduction in Fine Particulate Air Pollution and Mortality.
American Journal of Respiratory and Critical Care Medicine. 173:667-
672. EPA-HQ-OAR-2009-0472-1661.

[[Page 25526]]

Table III.H.7-2--Unquantified and Non-Monetized Potential Effects
------------------------------------------------------------------------
Effects not included in analysis--changes
Pollutant/effects in:
------------------------------------------------------------------------
Ozone Health \a\............. Chronic respiratory damage \b\.
Premature aging of the lungs \b\.
Non-asthma respiratory emergency room
visits.
Exposure to UVb (+/-) \e\.
Ozone Welfare................ Yields for
--commercial forests.
--some fruits and vegetables.
--non-commercial crops.
Damage to urban ornamental plants.
Impacts on recreational demand from
damaged forest aesthetics.
Ecosystem functions.
Exposure to UVb (+/-) \e\.
PM Health \c\................ Premature mortality--short term exposures
\d\.
Low birth weight.
Pulmonary function.
Chronic respiratory diseases other than
chronic bronchitis.
Non-asthma respiratory emergency room
visits.
Exposure to UVb (+/-) \e\.
PM Welfare................... Residential and recreational visibility
in non-Class I areas.
Soiling and materials damage.
Damage to ecosystem functions.
Exposure to UVb (+/-) \e\.
Nitrogen and Sulfate Commercial forests due to acidic sulfate
Deposition Welfare. and nitrate deposition.
Commercial freshwater fishing due to
acidic deposition.
Recreation in terrestrial ecosystems due
to acidic deposition.
Existence values for currently healthy
ecosystems.
Commercial fishing, agriculture, and
forests due to nitrogen deposition.
Recreation in estuarine ecosystems due to
nitrogen deposition.
Ecosystem functions.
Passive fertilization.
CO Health.................... Behavioral effects.
HC/Toxics Health \f\......... Cancer (benzene, 1,3-butadiene,
formaldehyde, acetaldehyde).
Anemia (benzene).
Disruption of production of blood
components (benzene).
Reduction in the number of blood
platelets (benzene).
Excessive bone marrow formation
(benzene).
Depression of lymphocyte counts
(benzene).
Reproductive and developmental effects
(1,3-butadiene).
Irritation of eyes and mucus membranes
(formaldehyde).
Respiratory irritation (formaldehyde).
Asthma attacks in asthmatics
(formaldehyde).
Asthma-like symptoms in non-asthmatics
(formaldehyde).
Irritation of the eyes, skin, and
respiratory tract (acetaldehyde).
Upper respiratory tract irritation and
congestion (acrolein).
HC/Toxics Welfare............ Direct toxic effects to animals.
Bioaccumulation in the food chain.
Damage to ecosystem function.
Odor.
------------------------------------------------------------------------
Notes:
\a\ The public health impact of biological responses such as increased
airway responsiveness to stimuli, inflammation in the lung, acute
inflammation and respiratory cell damage, and increased susceptibility
to respiratory infection are likely partially represented by our
quantified endpoints.
\b\ The public health impact of effects such as chronic respiratory
damage and premature aging of the lungs may be partially represented
by quantified endpoints such as hospital admissions or premature
mortality, but a number of other related health impacts, such as
doctor visits and decreased athletic performance, remain unquantified.
\c\ In addition to primary economic endpoints, there are a number of
biological responses that have been associated with PM health effects
including morphological changes and altered host defense mechanisms.
The public health impact of these biological responses may be partly
represented by our quantified endpoints.
\d\ While some of the effects of short-term exposures are likely to be
captured in the estimates, there may be premature mortality due to
short-term exposure to PM not captured in the cohort studies used in
this analysis. However, the PM mortality results derived from the
expert elicitation do take into account premature mortality effects of
short term exposures.
\e\ May result in benefits or disbenefits.
\f\ Many of the key hydrocarbons related to this rule are also hazardous
air pollutants listed in the CAA.

While there will be impacts associated with air toxic pollutant
emission changes that result from the final rule, we do not attempt to
monetize those impacts. This is primarily because currently available
tools and methods to assess air toxics risk from mobile sources at the
national scale are not adequate for extrapolation to incidence
estimations or benefits assessment. The best suite of tools and methods
currently available for assessment at the national scale are those used
in the National-Scale Air Toxics Assessment (NATA). The EPA Science
Advisory Board specifically commented in their review of the 1996 NATA
that these tools were not yet

[[Page 25527]]

ready for use in a national-scale benefits analysis, because they did
not consider the full distribution of exposure and risk, or address
sub-chronic health effects.\479\ While EPA has since improved the
tools, there remain critical limitations for estimating incidence and
assessing benefits of reducing mobile source air toxics. EPA continues
to work to address these limitations; however, we did not have the
methods and tools available for national-scale application in time for
the analysis of the final rule.\480\
---------------------------------------------------------------------------

\479\ Science Advisory Board. 2001. NATA--Evaluating the
National-Scale Air Toxics Assessment for 1996--an SAB Advisory.
http://www.epa.gov/ttn/atw/sab/sabrev.html. EPA-HQ-OAR-2009-0472-
0244.
\480\ In April 2009, EPA hosted a workshop on estimating the
benefits or reducing hazardous air pollutants. This workshop built
upon the work accomplished in the June 2000 Science Advisory Board/
EPA Workshop on the Benefits of Reductions in Exposure to Hazardous
Air Pollutants, which generated thoughtful discussion on approaches
to estimating human health benefits from reductions in air toxics
exposure, but no consensus was reached on methods that could be
implemented in the near term for a broad selection of air toxics.
Please visit http://epa.gov/air/toxicair/2009workshop.html. for more
information about the workshop and its associated materials.
---------------------------------------------------------------------------

EPA is also unaware of specific information identifying any effects
on listed endangered species from the small fluctuations in pollutant
concentrations associated with this rule (see Section III.G.5).
Furthermore, our current modeling tools are not designed to trace
fluctuations in ambient concentration levels to potential impacts on
particular endangered species.
i. Quantified Human Health Impacts
Tables III.H.7-3 and III.H.7-4 present the annual PM2.5
and ozone health impacts in the 48 contiguous U.S. states associated
with the final rule for 2030. For each endpoint presented in Tables
III.H.7-3 and III.H.7-4, we provide both the mean estimate and the 90%
confidence interval.
Using EPA's preferred estimates, based on the American Cancer
Society (ACS) and Six-Cities studies and no threshold assumption in the
model of mortality, we estimate that the final rule will result in
between 60 and 150 cases of avoided PM2.5-related premature
deaths annually in 2030. As a sensitivity analysis, when the range of
expert opinion is used, we estimate between 22 and 200 fewer premature
mortalities in 2030 (see Table 7.7 in the RIA that accompanies this
rule). For ozone-related premature mortality in 2030, we estimate a
range of between 4 to 18 additional premature mortalities related to
the ethanol production volumes mandated by the recently finalized RFS2
rule \481\ (and reflected in the air quality modeling for this rule),
but are not due to the final standards themselves.
---------------------------------------------------------------------------

\481\ EPA 2010, Renewable Fuel Standard Program (RFS2)
Regulatory Impact Analysis. EPA-420-R-10-006. February 2010. Docket
EPA-HQ-OAR-2009-0472-11332. EPA-HQ-OAR-2009-0472-11332. See also 75
FR 14670, March 26, 2010.

Table III.H.7-3--Estimated PM2.5-Related Health Impacts a
------------------------------------------------------------------------
2030 Annual reduction in
Health effect incidence (5th%-
95th%ile)
------------------------------------------------------------------------
Premature Mortality--Derived from
epidemiology literature: \b\
Adult, age 30+, ACS Cohort Study (Pope et 60 (23-96)
al., 2002).
Adult, age 25+, Six-Cities Study (Laden 150 (83-220)
et al., 2006).
Infant, age <1 year (Woodruff et al., 0 (0-1)
1997).
Chronic bronchitis (adult, age 26 and over).. 42 (8-77)
Non-fatal myocardial infarction (adult, age 100 (38-170)
18 and over).
Hospital admissions--respiratory (all ages) 13 (7-20)
\c\.
Hospital admissions--cardiovascular (adults, 32 (23-38)
age >18) \d\.
Emergency room visits for asthma (age 18 42 (25-59)
years and younger).
Acute bronchitis (children, age 8-12)........ 95 (0-190)
Lower respiratory symptoms (children, age 7- 1,100 (540-1,700)
14).
Upper respiratory symptoms (asthmatic 850 (270-1,400)
children, age 9-18).
Asthma exacerbation (asthmatic children, age 1,000 (120-2,900)
6-18).
Work loss days............................... 7,600 (6,600-8,500)
Minor restricted activity days (adults age 18- 45,000 (38,000-52,000)
65).
------------------------------------------------------------------------
Notes:

\a\ Incidence is rounded to two significant digits. Estimates represent
incidence within the 48 contiguous United States.
\b\ PM-related adult mortality based upon the American Cancer Society
(ACS) Cohort Study (Pope et al., 2002) and the Six-Cities Study (Laden
et al., 2006). Note that these are two alternative estimates of adult
mortality and should not be summed. PM-related infant mortality based
upon a study by Woodruff, Grillo, and Schoendorf (1997).\482\
\c\ Respiratory hospital admissions for PM include admissions for
chronic obstructive pulmonary disease (COPD), pneumonia and asthma.
\d\ Cardiovascular hospital admissions for PM include total
cardiovascular and subcategories for ischemic heart disease,
dysrhythmias, and heart failure.

---------------------------------------------------------------------------

\482\ Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 1997.
``The Relationship Between Selected Causes of Postneonatal Infant
Mortality and Particulate Air Pollution in the United States.''
Environmental Health Perspectives 105(6):608-612. EPA-HQ-OAR-2009-
0472-0382.

Table III.H.7-4--Estimated Ozone-Related Health Impacts a
------------------------------------------------------------------------
2030 Annual reduction in
Health effect incidence (5th%-
95th%ile)
------------------------------------------------------------------------
Premature Mortality, All ages \b\
Multi-City Analyses:
Bell et al. (2004)--Non-accidental....... -4 (-8-0)
Huang et al. (2005)--Cardiopulmonary..... -7 (-14-1)

[[Page 25528]]

Schwartz (2005)--Non-accidental.......... -6 (-13-1)
Meta-analyses:
Bell et al. (2005)--All cause............ -13 (-24--2)
Ito et al. (2005)--Non-accidental........ -18 (-30--6)
Levy et al. (2005)--All cause............ -18 (-28--9)
Hospital admissions--respiratory causes -38 (-86--6)
(adult, 65 and older) \c\.
Hospital admissions--respiratory causes -6 (-13-1)
(children, under 2).
Emergency room visit for asthma (all ages)... -16 (-51-8)
Minor restricted activity days (adults, age -18,000 (-40,000-3,700)
18-65).
School absence days.......................... -7,700 (-16,000-1,200)
------------------------------------------------------------------------
Notes:

\a\ Negatives indicate a disbenefit, or an increase in health effect
incidence. Incidence is rounded to two significant digits. Estimates
represent incidence within the 48 contiguous U.S.
\b\ Estimates of ozone-related premature mortality are based upon
incidence estimates derived from several alternative studies: Bell et
al. (2004); Huang et al. (2005); Schwartz (2005); Bell et al. (2005);
Ito et al. (2005); Levy et al. (2005). The estimates of ozone-related
premature mortality should therefore not be summed.
\c\ Respiratory hospital admissions for ozone include admissions for all
respiratory causes and subcategories for COPD and pneumonia.

ii. Monetized Benefits
Table III.H.7-5 presents the estimated monetary value of changes in
the incidence of ozone and PM2.5-related health effects. All
monetized estimates are stated in 2007$. These estimates account for
growth in real gross domestic product (GDP) per capita between the
present and 2030. Our estimate of total monetized benefits in 2030 for
the final rule, using the ACS and Six-Cities PM mortality studies and
the range of ozone mortality assumptions, is between $380 and $1,300
million, assuming a 3 percent discount rate, or between $330 and $1,200
million, assuming a 7 percent discount rate. As the results indicate,
total benefits are driven primarily by the reduction in
PM2.5-related premature fatalities each year.

Table III.H.7-5--Estimated Monetary Value of Changes in Incidence of Health and Welfare Effects
[In Millions of 2007$] a b
----------------------------------------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------
PM2.5-related health effect 2030
(5th and 95th%ile)
----------------------------------------------------------------------------------------------------------------
Premature Mortality--Derived from Adult, age 30+ --ACS study
Epidemiology Studies c d. (Pope et al., 2002)
3% discount rate............. $510 ($70-$1,300)
7% discount rate............. $460 ($63-$1,200)
Adult, age 25+ --Six-Cities
study (Laden et al., 2006)
3% discount rate............. $1,300 ($190-$3,300)
7% discount rate............. $1,200 ($180-$3,000)
Infant Mortality, <1 year-- $1.8 ($0-$7.0)
(Woodruff et al. 1997).
----------------------------------------------------------------------------------------------------------------
Chronic bronchitis (adults, 26 and over)............................ $22 ($1.9-$77)
Non-fatal acute myocardial infarctions
3% discount rate................................................ $14 ($3.9-$35)
7% discount rate................................................ $14 ($3.6-$35)
Hospital admissions for respiratory causes.......................... $0.20 ($0.01-$0.29)
Hospital admissions for cardiovascular causes....................... $0.91 ($0.58-$1.3)
Emergency room visits for asthma.................................... $0.016 ($0.009-$0.024)
Acute bronchitis (children, age 8-12)............................... $0.007 ($0-$0.018)
Lower respiratory symptoms (children, 7-14)......................... $0.022 ($0.009-$0.043)
Upper respiratory symptoms (asthma, 9-11)........................... $0.027 ($0.008-$0.061)
Asthma exacerbations................................................ $0.058 ($0.006-$0.17)
Work loss days...................................................... $1.2 ($1.0-$1.3)
Minor restricted-activity days (MRADs).............................. $2.9 ($1.7-$4.2)
----------------------------------------------------------------------------------------------------------------
Ozone-related Health Effect
----------------------------------------------------------------------------------------------------------------
Premature Mortality, All ages-- Bell et al., 2004............ -$38 (-$110-$4.2)
Derived from Multi-city analyses.
Huang et al., 2005........... -$62 (-$180-$4.7)
Schwartz, 2005............... -$58 (-$170-$8.8)
Premature Mortality, All ages-- Bell et al., 2005............ -$120 (-$330--$7.9)
Derived from Meta-analyses.
Ito et al., 2005............. -$170 (-$430--$19)
Levy et al., 2005............ -$170 (-$410--$21)
----------------------------------------------------------------------------------------------------------------
Hospital admissions--respiratory causes (adult, 65 and older)....... -$0.92 (-$2.1-$0.27)

[[Page 25529]]

PM2.5-related health effect 2030
(5th and 95th%ile)
----------------------------------------------------------------------------------------------------------------
Hospital admissions--respiratory causes (children, under 2)......... -$.21 (-$.45-$0.031)
Emergency room visit for asthma (all ages).......................... -$0.006 (-$0.018-$0.003)
Minor restricted activity days (adults, age 18-65).................. -$1.2 (-$2.7-$0.25)
School absence days................................................. -$0.71 (-$1.4-$0.11)
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ Negatives indicate a disbenefit, or an increase in health effect incidence. Monetary benefits are rounded to
two significant digits for ease of presentation and computation. PM and ozone benefits are nationwide.
\b\ Monetary benefits adjusted to account for growth in real GDP per capita between 1990 and the analysis year
(2030).
\c\ Valuation assumes discounting over the SAB recommended 20 year segmented lag structure. Results reflect the
use of 3 percent and 7 percent discount rates consistent with EPA and OMB guidelines for preparing economic
analyses.

iii. What are the limitations of the benefits analysis?
Every benefit-cost analysis examining the potential effects of a
change in environmental protection requirements is limited to some
extent by data gaps, limitations in model capabilities (such as
geographic coverage), and uncertainties in the underlying scientific
and economic studies used to configure the benefit and cost models.
Limitations of the scientific literature often result in the inability
to estimate quantitative changes in health and environmental effects,
such as potential increases in premature mortality associated with
increased exposure to carbon monoxide. Deficiencies in the economics
literature often result in the inability to assign economic values even
to those health and environmental outcomes which can be quantified.
These general uncertainties in the underlying scientific and economics
literature, which can lead to valuations that are higher or lower, are
discussed in detail in the RIA and its supporting references. Key
uncertainties that have a bearing on the results of the benefit-cost
analysis of the final rule include the following:
The exclusion of potentially significant and unquantified
benefit categories (such as health, odor, and ecological impacts of air
toxics, ozone, and PM);
Errors in measurement and projection for variables such as
population growth;
Uncertainties in the estimation of future year emissions
inventories and air quality;
Uncertainty in the estimated relationships of health and
welfare effects to changes in pollutant concentrations including the
shape of the C-R function, the size of the effect estimates, and the
relative toxicity of the many components of the PM mixture;
Uncertainties in exposure estimation; and
Uncertainties associated with the effect of potential
future actions to limit emissions.
As Table III.H.7-5 indicates, total benefits are driven primarily
by the reduction in PM2.5-related premature mortalities each
year. Some key assumptions underlying the premature mortality estimates
include the following, which may also contribute to uncertainty:
Inhalation of fine particles is causally associated with
premature death at concentrations near those experienced by most
Americans on a daily basis. Although biological mechanisms for this
effect have not yet been completely established, the weight of the
available epidemiological, toxicological, and experimental evidence
supports an assumption of causality. The impacts of including a
probabilistic representation of causality were explored in the expert
elicitation-based results of the PM NAAQS RIA.
All fine particles, regardless of their chemical
composition, are equally potent in causing premature mortality. This is
an important assumption, because PM produced via transported precursors
emitted from engines may differ significantly from PM precursors
released from electric generating units and other industrial sources.
However, no clear scientific grounds exist for supporting differential
effects estimates by particle type.
The C-R function for fine particles is approximately
linear within the range of ambient concentrations under consideration.
Thus, the estimates include health benefits from reducing fine
particles in areas with varied concentrations of PM, including both
regions that may be in attainment with PM2.5 standards and
those that are at risk of not meeting the standards.
There is uncertainty in the magnitude of the association
between ozone and premature mortality. The range of ozone impacts
associated with the final rule is estimated based on the risk of
several sources of ozone-related mortality effect estimates. In a
recent report on the estimation of ozone-related premature mortality
published by the National Research Council, a panel of experts and
reviewers concluded that short-term exposure to ambient ozone is likely
to contribute to premature deaths and that ozone-related mortality
should be included in estimates of the health benefits of reducing
ozone exposure.\483\ EPA has requested advice from the National Academy
of Sciences on how best to quantify uncertainty in the relationship
between ozone exposure and premature mortality in the context of
quantifying benefits.
---------------------------------------------------------------------------

\483\ National Research Council (NRC), 2008. Estimating
Mortality Risk Reduction and Economic Benefits from Controlling
Ozone Air Pollution. The National Academies Press: Washington, DC.
EPA-HQ-OAR-2009-0472-0322.
---------------------------------------------------------------------------

Acknowledging omissions and uncertainties, we present a best
estimate of the total benefits based on our interpretation of the best
available scientific literature and methods supported by EPA's
technical peer review panel, the Science Advisory Board's Health
Effects Subcommittee (SAB-HES). The National Academies of Science (NRC,
2002) has also reviewed EPA's methodology for analyzing the health
benefits of measures taken to reduce air pollution. EPA addressed many
of these comments in the analysis of the final PM
NAAQS.^{484 485} This

[[Page 25530]]

analysis incorporates this most recent work to the extent possible.
---------------------------------------------------------------------------

\484\ National Research Council (NRC). 2002. Estimating the
Public Health Benefits of Proposed Air Pollution Regulations. The
National Academies Press: Washington, DC.
\485\ U.S. Environmental Protection Agency. October 2006. Final
Regulatory Impact Analysis (RIA) for the National Ambient Air
Quality Standards for Particulate Matter. Prepared by: Office of Air
and Radiation. Available at http://www.epa.gov/ttn/ecas/ria.html.
EPA-HQ-OAR-2009-0472-0240.
---------------------------------------------------------------------------

b. PM-Related Monetized Benefits of the Model Year (MY) Analysis
As described in Section III.G, the final standards will reduce
emissions of several criteria and toxic pollutants and precursors. In
the MY analysis, EPA estimates the economic value of the human health
benefits associated with reducing PM2.5 exposure. Due to
analytical limitations, this analysis does not estimate benefits
related to other criteria pollutants (such as ozone, NO2 or
SO2) or toxics pollutants, nor does it monetize all of the
potential health and welfare effects associated with PM2.5.
The MY analysis uses a ``benefit-per-ton'' method to estimate a
selected suite of PM2.5-related health benefits described
below. These PM2.5 benefit-per-ton estimates provide the
total monetized human health benefits (the sum of premature mortality
and premature morbidity) of reducing one ton of directly emitted
PM2.5, or its precursors (such as NOX,
SOX, and VOCs), from a specified source. Ideally, the human
health benefits associated with the MY analysis would be estimated
based on changes in ambient PM2.5 as determined by full-
scale air quality modeling. However, this modeling was not possible in
the timeframe for the final rule.
The dollar-per-ton estimates used in this analysis are provided in
Table III.H.7-6. In the summary of costs and benefits, Section III.H.10
of this preamble, EPA presents the monetized value of PM-related
improvements associated with the rule.

Table III.H.7-6--Benefits-per-Ton Values (2007$) Derived Using the ACS Cohort Study for PM-Related Premature
Mortality (Pope et al., 2002) a
----------------------------------------------------------------------------------------------------------------
All sources \d\ Stationary (non-EGU) Mobile sources
-------------------------- sources -------------------------
Year \c\ --------------------------
SOX VOC Direct NOX Direct
NOX PM2.5 PM2.5
----------------------------------------------------------------------------------------------------------------
Estimated Using a 3 Percent Discount Rate \b\
----------------------------------------------------------------------------------------------------------------
2015.............................. $28,000 $1,200 $4,700 $220,000 $4,900 $270,000
2020.............................. 31,000 1,300 5,100 240,000 5,300 290,000
2030.............................. 36,000 1,500 6,100 280,000 6,400 350,000
2040.............................. 43,000 1,800 7,200 330,000 7,600 420,000
----------------------------------------------------------------------------------------------------------------
Estimated Using a 7 Percent Discount Rate \b\
----------------------------------------------------------------------------------------------------------------
2015.............................. 26,000 1,100 4,200 200,000 4,400 240,000
2020.............................. 28,000 1,200 4,600 220,000 4,800 270,000
2030.............................. 33,000 1,400 5,500 250,000 5,800 320,000
2040.............................. 39,000 1,600 6,600 300,000 6,900 380,000
----------------------------------------------------------------------------------------------------------------
\a\ The benefit-per-ton estimates presented in this table are based on an estimate of premature mortality
derived from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six-Cities
study (Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
\b\ The benefit-per-ton estimates presented in this table assume either a 3 percent or 7 percent discount rate
in the valuation of premature mortality to account for a twenty-year segmented cessation lag.
\c\ Benefit-per-ton values were estimated for the years 2015, 2020, and 2030. For 2040, EPA and NHTSA
extrapolated exponentially based on the growth between 2020 and 2030.
\d\ Note that the benefit-per-ton value for SOX is based on the value for Stationary (Non-EGU) sources; no SOX
value was estimated for mobile sources. The benefit-per-ton value for VOCs was estimated across all sources.

The benefit per-ton technique has been used in previous analyses,
including EPA's recent Ozone National Ambient Air Quality Standards
(NAAQS) RIA,\486\ the proposed Portland Cement National Emissions
Standards for Hazardous Air Pollutants (NESHAP) RIA,\487\ and the final
NO2 NAAQS (U.S. EPA, 2009b).\488\ Table III.H.7-7 shows the
quantified and unquantified PM2.5-related co-benefits
captured in those benefit-per-ton estimates.
---------------------------------------------------------------------------

\486\ U.S. Environmental Protection Agency (U.S. EPA). 2008.
Regulatory Impact Analysis, 2008 National Ambient Air Quality
Standards for Ground-level Ozone, Chapter 6. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. March. Available
at http://www.epa.gov/ttn/ecas/regdata/RIAs/6-ozoneriachapter6.pdf.
Accessed March 15, 2010. EPA-HQ-OAR-2009-0472-0108.
\487\ U.S. Environmental Protection Agency (U.S. EPA). 2009.
Regulatory Impact Analysis: National Emission Standards for
Hazardous Air Pollutants from the Portland Cement Manufacturing
Industry. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. April. Available on the Internet at http://
www.epa.gov/ttn/ecas/regdata/RIAs/portlandcementria_4-20-09.pdf.
Accessed March 15, 2010. EPA-HQ-OAR-2009-0472-0241.
\488\ U.S. Environmental Protection Agency (U.S. EPA). 2010.
Final NO2 NAAQS Regulatory Impact Analysis (RIA). Office of Air
Quality Planning and Standards, Research Triangle Park, NC. April.
Available on the Internet at http://www.epa.gov/ttn/ecas/regdata/
RIAs/FinalNO2RIAfulldocument.pdf. Accessed March 15, 2010. EPA-HQ-
OAR-2009-0472-0237.

Table III.H.7-7--Human Health and Welfare Effects of PM2.5
------------------------------------------------------------------------
Quantified and
Pollutant/effect monetized in primary Unquantified effects
estimates changes in
------------------------------------------------------------------------
PM2.5.................. Adult premature Subchronic bronchitis
mortality cases.
Bronchitis: chronic and Low birth weight.
acute
Hospital admissions: Pulmonary function.
respiratory and
cardiovascular.
Emergency room visits Chronic respiratory
for asthma diseases other than
chronic bronchitis.
Nonfatal heart attacks Non-asthma respiratory
(myocardial emergency room
infarction). visits.
Lower and upper Visibility.
respiratory illness

[[Page 25531]]

Minor restricted- Household soiling.
activity days
Work loss days
Asthma exacerbations
(asthmatic population)
Infant mortality
------------------------------------------------------------------------

Consistent with the NO2 NAAQS,\489\ the benefits
estimates utilize the concentration-response functions as reported in
the epidemiology literature. To calculate the total monetized impacts
associated with quantified health impacts, EPA applies values derived
from a number of sources. For premature mortality, EPA applies a value
of a statistical life (VSL) derived from the mortality valuation
literature. For certain health impacts, such as chronic bronchitis and
a number of respiratory-related ailments, EPA applies willingness-to-
pay estimates derived from the valuation literature. For the remaining
health impacts, EPA applies values derived from current cost-of-illness
and/or wage estimates.
---------------------------------------------------------------------------

\489\ Although we summarize the main issues in this chapter, we
encourage interested readers to see the benefits chapter of the
final NO2 NAAQS for a more detailed description of recent
changes to the PM benefits presentation and preference for the no-
threshold model.
---------------------------------------------------------------------------

Readers interested in reviewing the complete methodology for
creating the benefit-per-ton estimates used in this analysis can
consult the Technical Support Document (TSD) \490\ accompanying the
recent final ozone NAAQS RIA. Readers can also refer to Fann et al.
(2009) \491\ for a detailed description of the benefit-per-ton
methodology.\492\ A more detailed description of the benefit-per-ton
estimates is also provided in the Joint TSD that accompanies this
rulemaking.
---------------------------------------------------------------------------

\490\ U.S. Environmental Protection Agency (U.S. EPA). 2008b.
Technical Support Document: Calculating Benefit per-Ton estimates,
Ozone NAAQS Docket EPA-HQ-OAR-2007-0225-0284. Office of Air
Quality Planning and Standards, Research Triangle Park, NC. March.
Available on the Internet at http://www.regulations.gov. EPA-HQ-OAR-
2009-0472-0228.
\491\ Fann, N. et al. (2009). The influence of location, source,
and emission type in estimates of the human health benefits of
reducing a ton of air pollution. Air Qual Atmos Health. Published
online: 09 June, 2009. EPA-HQ-OAR-2009-0472-0229.
\492\ The values included in this report are different from
those presented in the article cited above. Benefits methods change
to reflect new information and evaluation of the science. Since
publication of the June 2009 article, EPA has made two significant
changes to its benefits methods: (1) We no longer assume that a
threshold exists in PM-related models of health impacts; and (2) We
have revised the Value of a Statistical Life to equal $6.3 million
(year 2000$), up from an estimate of $5.5 million (year 2000$) used
in the June 2009 report. Please refer to the following Web site for
updates to the dollar-per-ton estimates: http://www.epa.gov/air/
benmap/bpt.html. EPA-HQ-OAR-2009-0472-0227.
---------------------------------------------------------------------------

As described in the documentation for the benefit per-ton estimates
cited above, national per-ton estimates were developed for selected
pollutant/source category combinations. The per-ton values calculated
therefore apply only to tons reduced from those specific pollutant/
source combinations (e.g., NO2 emitted from mobile sources;
direct PM emitted from stationary sources). Our estimate of
PM2.5 benefits is therefore based on the total direct
PM2.5 and PM-related precursor emissions controlled by
sector and multiplied by each per-ton value.
The benefit-per-ton estimates are subject to a number of
assumptions and uncertainties.
Dollar-per-ton estimates do not reflect local variability
in population density, meteorology, exposure, baseline health incidence
rates, or other local factors that might lead to an overestimate or
underestimate of the actual benefits of controlling fine particulates.
In Section III.G, we describe the full-scale air quality modeling
conducted for the 2030 calendar year analysis in an effort to capture
this variability.
There are several health benefits categories that EPA was
unable to quantify in the MY analysis due to limitations associated
with using benefits-per-ton estimates, several of which could be
substantial. Because NOX and VOC emissions are also
precursors to ozone, changes in NOX and VOC would also
impact ozone formation and the health effects associated with ozone
exposure. Benefits-per-ton estimates do not exist for ozone, however,
due to issues associated with the complexity of the atmospheric air
chemistry and nonlinearities associated with ozone formation. The PM-
related benefits-per-ton estimates also do not include any human
welfare or ecological benefits. Please refer to Chapter 7 of the RIA
that accompanies this rule for a description of the quantification and
monetization of health impacts for the CY analysis and a description of
the unquantified co-pollutant benefits associated with this rulemaking.
The benefit-per-ton estimates used in this analysis
incorporate projections of key variables, including atmospheric
conditions, source level emissions, population, health baselines and
incomes, technology. These projections introduce some uncertainties to
the benefit per ton estimates.
As described above, using the benefit-per-ton value
derived from the ACS study (Pope et al., 2002) alone provides an
incomplete characterization of PM2.5 benefits. When placed
in the context of the Expert Elicitation results, this estimate falls
toward the lower end of the distribution. By contrast, the estimated
PM2.5 benefits using the coefficient reported by Laden in
that author's reanalysis of the Harvard Six-Cities cohort fall toward
the upper end of the Expert Elicitation distribution results.
As mentioned above, emissions changes and benefits-per-ton
estimates alone are not a good indication of local or regional air
quality and health impacts, as there may be localized impacts
associated with this rulemaking. Additionally, the atmospheric
chemistry related to ambient concentrations of PM2.5, ozone
and air toxics is very complex. Full-scale photochemical modeling is
therefore necessary to provide the needed spatial and temporal detail
to more completely and accurately estimate the changes in ambient
levels of these pollutants and their associated health and welfare
impacts. Timing and resource constraints precluded EPA from conducting
full-scale photochemical air quality modeling for the MY analysis. We
have, however, conducted national-scale air quality modeling for the CY
analysis to analyze the impacts of the standards on PM2.5,
ozone, and selected air toxics.
8. Energy Security Impacts
This rule to reduce GHG emissions in light-duty vehicles results in
improved fuel efficiency which, in turn, helps to reduce U.S. petroleum
imports. A reduction of U.S. petroleum imports reduces both financial
and strategic risks caused by potential sudden disruptions in the
supply of imported petroleum to the U.S. This reduction in

[[Page 25532]]

risk is a measure of improved U.S. energy security. This section
summarizes our estimate of the monetary value of the energy security
benefits of the GHG vehicle standards against the reference case by
estimating the impact of the expanded use of lower-GHG vehicle
technologies on U.S. oil imports and avoided U.S. oil import
expenditures. Additional discussion of this issue can be found in
Chapter 5.1 of EPA's RIA and Section 4.2.8 of the TSD.
a. Implications of Reduced Petroleum Use on U.S. Imports
In 2008, U.S. petroleum import expenditures represented 21 percent
of total U.S. imports of all goods and services.\493\ In 2008, the U.S.
imported 66 percent of the petroleum it consumed, and the
transportation sector accounted for 70 percent of total U.S. petroleum
consumption. This compares to approximately 37 percent of petroleum
from imports and 55 percent of consumption from petroleum in the
transportation sector in 1975.\494\ It is clear that petroleum imports
have a significant impact on the U.S. economy. Requiring lower-GHG
vehicle technology in the U.S. is expected to lower U.S. petroleum
imports.
---------------------------------------------------------------------------

\493\ Source: U.S. Bureau of Economic Analysis, U.S.
International Transactions Accounts Data, as shown on June 24, 2009.
\494\ Source: U.S. Department of Energy, Annual Energy Review
2008, Report No. DOE/EIA-0384(2008), Tables 5.1 and 5.13c, June 26,
2009.
---------------------------------------------------------------------------

b. Energy Security Implications
In order to understand the energy security implications of reducing
U.S. petroleum imports, EPA worked with Oak Ridge National Laboratory
(ORNL), which has developed approaches for evaluating the economic
costs and energy security implications of oil use. The energy security
estimates provided below are based upon a methodology developed in a
peer-reviewed study entitled ``The Energy Security Benefits of Reduced
Oil Use, 2006-2015,'' completed in March 2008. This study is included
as part of the docket for this rulemaking.^{495 496}
---------------------------------------------------------------------------

\495\ Leiby, Paul N. ``Estimating the Energy Security Benefits
of Reduced U.S. Oil Imports'' Oak Ridge National Laboratory, ORNL/
TM-2007/028, Final Report, 2008. (Docket EPA-HQ-OAR-2009-0472).
\496\ The ORNL study ``The Energy Security Benefits of Reduced
Oil Use, 2006-2015,'' completed in March 2008, is an update version
of the approach used for estimating the energy security benefits of
U.S. oil import reductions developed in an ORNL 1997 Report by
Leiby, Paul N., Donald W. Jones, T. Randall Curlee, and Russell Lee,
entitled ``Oil Imports: An Assessment of Benefits and Costs.''
(Docket EPA-HQ-OAR-2009-0472).
---------------------------------------------------------------------------

When conducting this analysis, ORNL considered the economic cost of
importing petroleum into the U.S. The economic cost of importing
petroleum into the U.S. is defined to include two components in
addition to the purchase price of petroleum itself. These are: (1) The
higher costs for oil imports resulting from the effect of increasing
U.S. import demand on the world oil price and on OPEC market power
(i.e., the ``demand'' or ``monopsony'' costs); and (2) the risk of
reductions in U.S. economic output and disruption of the U.S. economy
caused by sudden disruptions in the supply of imported petroleum to the
U.S. (i.e., macroeconomic disruption/adjustment costs). Maintaining a
U.S. military presence to help secure stable oil supply from
potentially vulnerable regions of the world was not included in this
analysis because its attribution to particular missions or activities
is hard to quantify.
One commenter on this rule felt that the magnitude of the economic
disruption portion of the energy security benefit may be too high. This
commenter cites a recent paper written by Stephen P.A. Brown and
Hillard G. Huntington, entitled ``Estimating U.S. Oil Security
Premiums'' (September 2009) as the basis for their comment. The Agency
reviewed this paper and found that it conducted a somewhat different
analysis than the one conducted by ORNL in support of this rule. The
Brown and Huntington paper focuses on policies and the energy security
implications of increasing U.S. demand for oil (or at least holding
U.S. oil consumption constant), while the ORNL analysis examines the
energy security implications of decreasing U.S. oil consumption and oil
imports. These asymmetrical analyses would be expected to yield
somewhat different energy security results.
However, even given the different scenarios considered, the Brown
and Huntington estimates are roughly similar to the ORNL estimates. For
example, for an increase in U.S. consumption that leads to an increase
in U.S. imports of oil, Brown and Huntington estimate a 2015 disruption
premium of $4.87 per barrel, with an uncertainty range from $1.03 to
$14.10 per barrel. The corresponding 2015 estimate for ORNL as the
result of a reduction in U.S. oil imports is $6.70 per barrel, with an
uncertainty range of $3.11 to $10.67 per barrel. Given that the two
studies analyze different scenarios, since the Brown and Huntington
disruption premiums are well within the uncertainty range of the ORNL
study, and given that the ORNL scenario matches the specific oil market
impacts anticipated from the rule while the Brown and Huntington paper
does not, the Agency has concluded that the ORNL disruption security
premium estimates are more applicable for analyzing this final rule.
In the energy security literature, the macroeconomic disruption
component of the energy security premium traditionally has included
both (1) increased payments for petroleum imports associated with a
rapid increase in world oil prices, and (2) the GDP losses and
adjustment costs that result from projected future oil price shocks.
One commenter suggested that the increased payments associated with
rapid increases in petroleum prices (i.e., price increases in a
disrupted market) represent transfers from U.S. oil consumers to
petroleum suppliers rather than real economic costs, and therefore,
should not be counted as a benefit.
This approach would represent a significant departure from how the
macroeconomic disruption costs associated with oil price shocks have
been quantified in the broader energy security literature, and the
Agencies believe it should be analyzed in more detail before being
applied in a regulatory context. In addition, the Agencies also believe
that there are compelling reasons to treat higher oil import costs
during oil supply disruptions differently than simple wealth transfers
that reflect the exercise of market power by petroleum sellers or
consumers. According to the OMB definition of a transfer: ``Benefit and
cost estimates should reflect real resource use. Transfer payments are
monetary payments from one group to another that do not affect total
resources available to society. * * * The net reduction in the total
surplus (consumer plus producer) is a real cost to society, but the
transfer from buyers to sellers resulting from a higher price is not a
real cost since the net reduction automatically accounts for the
transfer from buyers to sellers.'' \497\ In other words, pure transfers
do not lead to changes in the allocation or consumption of economic
resources, whereas changes in the resource allocation or use produce
real economic costs or benefits.
---------------------------------------------------------------------------

\497\ OMB Circular A-4, September 17, 2003. See http://
www.whitehouse.gov/omb/assets/omb/circulars/a004/a-4.pdf.
---------------------------------------------------------------------------

While price increases during oil price disruptions can result in
large transfers of wealth, they also result in a combination of real
resource shortages, costly short-run shifts in energy supply,
behavioral and demand adjustments by energy users, and other response
costs. Unlike pure transfers, the root cause of

[[Page 25533]]

the disruption price increase is a real resource supply reduction due,
for example, to disaster or war. Regions where supplies are disrupted
(i.e., the U.S.) suffer very high costs. Businesses' and households'
emergency responses to supply disruptions and rapid price increases are
likely to consume some real economic resources, in addition to causing
financial losses to the U.S. economy that are matched by offsetting
gains elsewhere in the global economy.
While households and businesses can reduce their petroleum
consumption, invest in fuel switching technologies, or use futures
markets to insulate themselves in advance against the potential costs
of rapid increases in oil prices, when deciding how extensively to do
so, they are unlikely to account for the effect of their petroleum
consumption on the magnitude of costs that supply interruptions and
accompanying price shocks impose on others. As a consequence, the U.S.
economy as a whole will not make sufficient use of these mechanisms to
insulate itself from the real costs of rapid increases in energy prices
and outlays that usually accompany oil supply interruptions.\498\
Therefore, the ORNL estimate of macroeconomic disruption and adjustment
costs that the Agencies use to value energy security benefits includes
the increased oil import costs stemming from oil price shocks that are
unanticipated and not internalized by advance actions of U.S. consumers
of petroleum products. The Agencies believe that, as the ORNL analysis
argues, the uninternalized oil import costs that occur during oil
supply interruptions represents a real cost associated with U.S.
petroleum consumption and imports, and that reducing its value by
lowering domestic petroleum consumption and imports thus represents a
real economic benefit from lower fuel consumption.
---------------------------------------------------------------------------

\498\ For a more complete discussion of the reasons why the oil
import cost component of the macroeconomic disruption and adjustment
costs includes some real costs and does not represent a pure
transfer, see Paul N. Leiby, Estimating the Energy Security Benefits
of Reduced U.S. Oil Imports: Final Report, ORNL-TM-2007-028, Oak
Ridge National Laboratory, March 14, 2008, pp. 21-25.
---------------------------------------------------------------------------

For this rule, ORNL estimated the energy security premium by
incorporating the oil price forecast of the Energy Information
Administration's 2009 Annual Energy Outlook (AEO) to its model. The
Agency considered, but rejected the option, of further updating this
analysis using the oil price estimates provided by the AEO 2010. Given
the broad uncertainty bands around oil price forecasts and the
relatively modest change in oil price forecasts between the AEO 2009
and AEO 2010, the Agency felt that updating to AEO 2010 oil prices
would not significantly change the results of this energy security
analysis. Finally, the EPA used its OMEGA model in conjunction with
ORNL's energy security premium estimates to develop the total energy
security benefits for a number of different years; please refer to
Table III.H.8-1 for this information for years 2015, 2020, 2030 and
2040,\499\ as well as a breakdown of the components of the energy
security premium for each of these years. The components of the energy
security premium and their values are discussed in detail in the Joint
TSD Chapter 4.
---------------------------------------------------------------------------

\499\ AEO 2009 forecasts energy market trends and values only to
2030. The energy security premium estimates post-2030 were assumed
to be the 2030 estimate.
---------------------------------------------------------------------------

Because the price of oil is determined globally, supply and demand
shocks anywhere in the world will have an adverse impact on the United
States (and on all other oil consuming countries). The total economic
costs of those shocks to the U.S. will depend on both U.S. petroleum
consumption and imports of petroleum and refined products. The analysis
relied upon to estimate energy security benefits from reducing U.S.
petroleum consumption estimates the value of energy security using the
estimated oil import premium, and is thus consistent with how much of
the energy security literature reports energy security impacts. Since
this rule is expected to have little impact on the U.S. supply of crude
petroleum, a reduction in U.S. fuel consumption is expected to be
reflected predominantly in reduced imports of petroleum and refined
fuel. The estimated energy security premium associated with a reduction
in U.S. petroleum consumption that leads to a reduction in imports
would likely be somewhat larger, due to diminished sensitivity of the
U.S. economy to oil supply shocks that would accompany the reduction in
oil consumption.
In addition, while the estimates of energy security externalities
used in this analysis depend on a combination of U.S. petroleum
consumption and imports, they have been expressed as per barrel of
petroleum imported into the U.S. The Agencies' analyses apply these
estimates to the reduction in U.S. imports of crude petroleum and
refined products that is projected to result from the rule in order to
determine the benefits that are likely to result from fuel savings and
the consequent reduction in imports. Thus, the estimates of energy
security externalities have been used in this analysis in a way that is
completely consistent with how they are defined and measured in the
ORNL analysis.

Table III.H.8-1--Energy Security Premium in 2015, 2020, 2030 and 2040 (2007$/Barrel)
----------------------------------------------------------------------------------------------------------------
Macroeconomic
Year (range) Monopsony disruption/adjustment Total mid-point
costs
----------------------------------------------------------------------------------------------------------------
2015................................. $11.79 ($4.26-$21.37).. $6.70 ($3.11-$10.67)... $18.49 ($9.80-$28.08)
2020................................. $12.31 ($4.46-$22.53).. $7.62 ($3.77-$12.46)... $19.94 ($10.58-$30.47)
2030................................. $10.57 ($3.84-18.94)... $8.12 ($3.90-$13.04)... $18.69 ($10.52-$27.89)
2040................................. $10.57 ($3.84-$18.94).. $8.12 ($3.90-$13.04)... $18.69 ($10.52-$27.89)
----------------------------------------------------------------------------------------------------------------

The literature on the energy security for the last two decades has
routinely combined the monopsony and the macroeconomic disruption
components when calculating the total value of the energy security
premium. However, in the context of using a global value for the Social
Cost of Carbon (SCC) the question arises: How should the energy
security premium be used when some benefits from the rule, such as the
benefits of reducing greenhouse gas emissions, are calculated using a
global value? Monopsony benefits represent avoided payments by the U.S.
to oil producers in foreign countries that result from a decrease in
the world oil price as the U.S. decreases its consumption of imported
oil. Although there is clearly a benefit to the U.S. when considered
from the domestic perspective, the decrease in price due to decreased
demand in the U.S. also represents a loss of income to oil-

[[Page 25534]]

producing countries. Given the redistributive nature of this effect, do
the negative effects on other countries ``net out'' the positive
impacts to the U.S.? If this is the case, then the monopsony portion of
the energy security premium should be excluded from the net benefits
calculation for the rule. OMB's Circular A-4 gives guidance in this
regard. Domestic pecuniary benefits (or transfers between buyers and
sellers) generally should not be included because they do not represent
real resource costs, though A-4 notes that transfers to the U.S. from
other countries may be counted as benefits as long as the analysis is
conducted from a U.S. perspective.
Energy security is broadly defined as protecting the U.S. economy
against circumstances that threaten significant short- and long-term
increases in energy costs. Energy security is inherently a domestic
benefit. Accordingly, it is possible to argue that the use of the
domestic monopsony benefit may not necessarily be in conflict with the
use of the global SCC, because the global SCC represents the benefits
against which the costs of our (i.e., the U.S.'s) domestic mitigation
efforts should be judged. In the final analysis, the Agency has
determined that using only the macroeconomic disruption component of
the energy security benefit is the appropriate metric for this rule.
At proposal, the Agency took the position that since a global
perspective was being taken with the use of the global SCC, that the
monopsony benefits ``net out'' and were a transfer. Two commenters felt
that the monopsony effect should be excluded from net benefits
calculations for the rule since it is a ``pecuniary'' externality or
does not represent an efficiency gain. One of the commenters suggested
that EPA instead conduct a distributional analysis of the monopsony
impacts of the final rule. The Agency disagrees that all pecuniary
externalities should necessarily be excluded from net benefits
calculations as a general rule. In this case considered here, the oil
market is non-competitive, and if the social decision-making unit of
interest is the U.S., there is an argument for accounting for the
monopsony premium to assess the excess transfer of wealth caused by the
exercise of cartel power outside of the U.S.
However, for the final rule, the Agency continues to take a global
perspective with respect to climate change by using the global SCC.
Therefore, the Agency did not count monopsony benefits since they ``net
out'' with losses to other countries outside the U.S. Since a global
perspective has been taken, a distributional analysis was not
undertaken for this final rule, since the losses to the losers (oil
producers that export oil to the U.S.) would equal the gains to the
winners (U.S. consumers of imported oil). As a result, the Agency has
included only the macroeconomic disruption portion of the energy
security benefits to monetize the total energy security benefits of
this rule. Hence, the total annual energy security benefits are derived
from the estimated reductions in U.S. imports of finished petroleum
products and crude oil using only the macroeconomic disruption/
adjustment portion of the energy security premium. These values are
shown in Table III.H.8-2.\500\ The reduced oil estimates were derived
from the OMEGA model, as explained in Section III.F of this preamble.
EPA used the same assumption that NHTSA used in its Corporate Average
Fuel Economy and CAFE Reform for MY 2008-2011 Light Trucks rule, which
assumed that each gallon of fuel saved reduces total U.S. imports of
crude oil or refined products by 0.95 gallons.\501\
---------------------------------------------------------------------------

\500\ Estimated reductions in U.S. imports of finished petroleum
products and crude oil are 95% of 89 million barrels (MMB) in 2015,
300 MMB in 2020, 590 MMB in 2030, and 778 MMB in 2040.
\501\ Preliminary Regulatory Impacts Analysis, April 2008. Based
on a detailed analysis of differences in fuel consumption, petroleum
imports, and imports of refined petroleum products among the
Reference Case, High Economic Growth, and Low Economic Growth
Scenarios presented in the Energy Information Administration's
Annual Energy Outlook 2007, NHTSA estimated that approximately 50
percent of the reduction in fuel consumption is likely to be
reflected in reduced U.S. imports of refined fuel, while the
remaining 50 percent would be expected to be reflected in reduced
domestic fuel refining. Of this latter figure, 90 percent is
anticipated to reduce U.S. imports of crude petroleum for use as a
refinery feedstock, while the remaining 10 percent is expected to
reduce U.S. domestic production of crude petroleum. Thus on balance,
each gallon of fuel saved is anticipated to reduce total U.S.
imports of crude petroleum or refined fuel by 0.95 gallons.

Table III.H.8-2--Total Annual Energy Security Benefits Using Only the
Macroeconomic Disruption/Adjustment Component of the Energy Security
Premium in 2015, 2020, 2030 and 2040
[Billions of 2007$]
------------------------------------------------------------------------
Year Benefits
------------------------------------------------------------------------
2015.................................................... $0.57
2020.................................................... $2.17
2030.................................................... $4.55
2040.................................................... $6.00
------------------------------------------------------------------------

9. Other Impacts
There are other impacts associated with the CO2
emissions standards and associated reduced fuel consumption that vary
with miles driven. Lower fuel consumption would, presumably, result in
fewer trips to the filling station to refuel and, thus, time saved. The
rebound effect, discussed in detail in Section III.H.4.c, produces
additional benefits to vehicle owners in the form of consumer surplus
from the increase in vehicle-miles driven, but may also increase the
societal costs associated with traffic congestion, motor vehicle
crashes, and noise. These effects are likely to be relatively small in
comparison to the value of fuel saved as a result of the standards, but
they are nevertheless important to include. Table III.H.9-1 summarizes
the other economic impacts. Please refer to Preamble Section II.F and
the Joint TSD that accompanies this rule for more information about
these impacts and how EPA and NHTSA use them in their analyses.
Note that for the estimated value of less frequent refueling
events, EPA's estimate is subject to a number of uncertainties which we
discuss in detail in Chapter 4.1.11 of the Joint TSD, and the actual
value could be higher or lower than the value presented here.
Specifically, the analysis makes three assumptions: (a) That
manufacturers will not adjust fuel tank capacities downward (from the
current average of 19.3 gallons) when they improve the fuel economy of
their vehicle models. (b) that the average fuel purchase (55 percent of
fuel tank capacity) is the typical fuel purchase. (c) that 100 percent
of all refueling is demand-based; i.e., that every gallon of fuel which
is saved would reduce the need to return to the refueling station. A
new research project is being planned by DOT which will include a
detailed study of refueling events, and which is expected to improve
upon these assumptions. These assumptions and the new DOT research
project are discussed in detail in Joint TSD Chapter 4.2.10.

[[Page 25535]]

Table III.H.9-1--Other Impacts Associated With the Light-Duty Vehicle GHG Program
[Millions of 2007 dollars]
----------------------------------------------------------------------------------------------------------------
2020 2030 2040 2050 NPV, 3% NPV, 7%
----------------------------------------------------------------------------------------------------------------
Value of Less Frequent Refueling.. $2,400 $4,800 $6,300 $8,000 $87,900 $40,100
Value of Increased Driving \a\.... 4,200 8,800 13,000 18,400 171,500 75,500
Accidents, Noise, Congestion...... -2,300 -4,600 -6,100 -7,800 -84,800 -38,600
----------------------------------------------------------------------------------------------------------------
\a\ Calculated using post-tax fuel prices.

10. Summary of Costs and Benefits
In this section, EPA presents a summary of costs, benefits, and net
benefits of the rule. Table III.H.10-1 shows the estimated annual
societal costs of the vehicle program for the indicated calendar years.
The table also shows the net present values of those costs for the
calendar years 2012-2050 using both a 3 percent and a 7 percent
discount rate. In this table, fuel savings are calculated using pre-tax
fuel prices.
Consumers are expected to receive the fuel savings presented here.
The cost estimates for the fuel-saving technology are based on designs
that will hold all vehicle attributes constant except fuel economy and
technology cost. This analysis also assumes that consumers will not
change the vehicles that they purchase. Automakers may redesign
vehicles as part of their compliance strategies. The redesigns should
be expected to make the vehicles more attractive to consumers, because
the ability to hold all other attributes constant means that the only
reason to change them is to make them more marketable to consumers. In
addition, consumers may choose to purchase different vehicles than they
would in the absence of this rule. These changes may affect the net
benefits that consumers receive from their vehicles. If consumers can
buy the same vehicle as before, except with increased price and fuel
economy, then the increase in vehicle price is the maximum loss in
welfare to the consumer, because compensating the increase in price
would leave her able to buy her previous vehicle with no change. If she
decides to purchase a different vehicle, or not to purchase a vehicle,
she would do so only if she were better off than buying her original
choice. Because of the unsettled state of the modeling of consumer
choices (discussed in Section III.H.1 and in RIA Section 8.1.2), this
analysis does not measure these effects. If the technology costs are
not sufficient to maintain other vehicle attributes, then it is
possible that automakers would be required to make less marketable
vehicles in order to comply with the rule; as a result, there may be an
additional loss in consumer welfare due to the rule. While EPA received
comments expressing concern over the possibility of these losses, there
were no specific losses identified.

Table III.H.10-1--Estimated Societal Costs of the Light-Duty Vehicle GHG Program
[Millions of 2007 dollars]
----------------------------------------------------------------------------------------------------------------
Social costs 2020 2030 2040 2050 NPV, 3% NPV, 7%
----------------------------------------------------------------------------------------------------------------
Vehicle Compliance Costs.......... $15,600 $15,800 $17,400 $19,000 $345,900 $191,900
Fuel Savings \a\.................. -35,700 -79,800 -119,300 -171,200 -1,545,600 -672,600
Quantified Annual Costs........... -20,100 -64,000 -101,900 -152,200 -1,199,700 -480,700
----------------------------------------------------------------------------------------------------------------
\a\ Calculated using pre-tax fuel prices.

Table III.H.10-2 presents estimated annual societal benefits for
the indicated calendar years. The table also shows the net present
values of those benefits for the calendar years 2012-2050 using both a
3 percent and a 7 percent discount rate. The table shows the benefits
of reduced CO2 emissions--and consequently the annual
quantified benefits (i.e., total benefits)--for each of four SCC values
considered by EPA. As discussed in the RIA Section 7.5, the IPCC Fourth
Assessment Report (2007) concluded that that the benefit estimates from
CO2 reductions are ``very likely'' underestimates. One of
the primary reasons is that models used to calculate SCC values do not
include information about impacts that have not been quantified.
In addition, these monetized GHG benefits exclude the value of
reductions in non-CO2 GHG emissions (HFC, CH4,
N2O) expected under this final rule. Although EPA has not
monetized the benefits of reductions in non-CO2 GHGs, the
value of these reductions should not be interpreted as zero. Rather,
the reductions in non-CO2 GHGs will contribute to this
rule's climate benefits, as explained in Section III.F. The SCC TSD
notes the difference between the social cost of non-CO2
emissions and SCC and specifies a goal to develop methods to value non-
CO2 emissions in future analyses.

Table III.H.10-2--Estimated Societal Benefits Associated With the Light-Duty Vehicle GHG Program
[Millions of 2007 dollars]
----------------------------------------------------------------------------------------------------------------
Benefits category 2020 2030 2040 2050 NPV, 3% \a\ NPV, 7% \a\
----------------------------------------------------------------------------------------------------------------
Reduced CO2 Emissions at
each assumed SCC value b c
Avg SCC at 5%........... $900 $2,700 $4,600 $7,200 $34,500 $34,500
Avg SCC at 3%........... 3,700 8,900 14,000 21,000 176,700 176,700
Avg SCC at 2.5%......... 5,800 14,000 21,000 30,000 299,600 299,600
95th percentile SCC at 11,000 27,000 43,000 62,000 538,500 538,500
3%.....................
Criteria Pollutant Benefits B 1,200-1,300 1,200-1,300 1,200-1,300 21,000 14,000
d e f g....................
Energy Security Impacts 2,200 4,500 6,000 7,600 81,900 36,900
(price shock)..............

[[Page 25536]]

Reduced Refueling........... 2,400 4,800 6,300 8,000 87,900 40,100
Value of Increased Driving h 4,200 8,800 13,000 18,400 171,500 75,500
Accidents, Noise, Congestion -2,300 -4,600 -6,100 -7,800 -84,800 -38,600
Quantified Annual Benefits
at each assumed SCC value b
c
Avg SCC at 5%........... 7,400 17,500 25,100 34,700 312,000 162,400
Avg SCC at 3%........... 10,200 23,700 34,500 48,500 454,200 304,600
Avg SCC at 2.5%......... 12,300 28,800 41,500 57,500 577,100 427,500
95th percentile SCC at 17,500 41,800 63,500 89,500 816,000 666,400
3%.....................
----------------------------------------------------------------------------------------------------------------
\a\ Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used
to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
\b\ Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
\c\ Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC
estimates range as follows: for Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $36-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also presents these SCC estimates.
\d\ Note that ``B'' indicates unquantified criteria pollutant benefits in the year 2020. For the final rule, we
only modeled the rule's PM2.5- and ozone-related impacts in the calendar year 2030. For the purposes of
estimating a stream of future-year criteria pollutant benefits, we assume that the benefits out to 2050 are
equal to, and no less than, those modeled in 2030 as reflected by the stream of estimated future emission
reductions. The NPV of criteria pollutant-related benefits should therefore be considered a conservative
estimate of the potential benefits associated with the final rule.
\e\ The benefits presented in this table include an estimate of PM-related premature mortality derived from
Laden et al., 2006, and the ozone-related premature mortality estimate derived from Bell et al., 2004. If the
benefit estimates were based on the ACS study of PM-related premature mortality (Pope et al., 2002) and the
Levy et al., 2005 study of ozone-related premature mortality, the values would be as much as 70% smaller.
\f\ The calendar year benefits presented in this table assume either a 3% discount rate in the valuation of PM-
related premature mortality ($1,300 million) or a 7% discount rate ($1,200 million) to account for a twenty-
year segmented cessation lag. Note that the benefits estimated using a 3% discount rate were used to calculate
the NPV using a 3% discount rate and the benefits estimated using a 7% discount rate were used to calculate
the NPV using a 7% discount rate. For benefits totals presented at each calendar year, we used the mid-point
of the criteria pollutant benefits range ($1,250).
\g\ Note that the co-pollutant impacts presented here do not include the full complement of endpoints that, if
quantified and monetized, would change the total monetized estimate of impacts. The full complement of human
health and welfare effects associated with PM and ozone remain unquantified because of current limitations in
methods or available data. We have not quantified a number of known or suspected health effects linked with
ozone and PM for which appropriate health impact functions are not available or which do not provide easily
interpretable outcomes (e.g., changes in heart rate variability). Additionally, we are unable to quantify a
number of known welfare effects, including reduced acid and particulate deposition damage to cultural
monuments and other materials, and environmental benefits due to reductions of impacts of eutrophication in
coastal areas.
\h\ Calculated using pre-tax fuel prices.

Table III.H.10-3 presents estimated annual net benefits for the
indicated calendar years. The table also shows the net present values
of those net benefits for the calendar years 2012-2050 using both a 3
percent and a 7 percent discount rate. The table includes the benefits
of reduced CO2 emissions (and consequently the annual net
benefits) for each of four SCC values considered by EPA. As noted
above, the benefit estimates from CO2 reductions are ``very
likely,'' according to the IPCC Fourth Assessment Report,
underestimates because, in part, models used to calculate SCC values do
not include information about impacts that have not been quantified.

Table III.H.10-3--Quantified Net Benefits Associated with the Light-Duty Vehicle GHG Program \a\
[Millions of 2007 dollars]
----------------------------------------------------------------------------------------------------------------
2020 2030 2040 2050 NPV, 3% b NPV, 7% b
----------------------------------------------------------------------------------------------------------------
Quantified Annual Costs..... -$20,100 -$64,000 -$101,900 -$152,200 -$1,199,700 -$480,700
----------------------------------------------------------------------------------------------------------------
Quantified Annual Benefits at each assumed SCC value c d
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 7,400 17,500 25,100 34,700 312,000 162,400
Avg SCC at 3%............... 10,200 23,700 34,500 48,500 454,200 304,600
Avg SCC at 2.5%............. 12,300 28,800 41,500 57,500 577,100 427,500
95th percentile SCC at 3%... 17,500 41,800 63,500 89,500 816,000 666,400
----------------------------------------------------------------------------------------------------------------
Quantified Net Benefits at each assumed SCC value c d
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 27,500 81,500 127,000 186,900 1,511,700 643,100
Avg SCC at 3%............... 30,300 87,700 136,400 200,700 1,653,900 785,300
Avg SCC at 2.5%............. 32,400 92,800 143,400 209,700 1,776,800 908,200

[[Page 25537]]

95th percentile SCC at 3%... 37,600 105,800 165,400 241,700 2,015,700 1,147,100
----------------------------------------------------------------------------------------------------------------
a Fuel impacts were calculated using pre-tax fuel prices.
b Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used
to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
c Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
d Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC
estimates range as follows: For Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $36-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also presents these SCC estimates.

EPA also conducted a separate analysis of the total benefits over
the model year lifetimes of the 2012 through 2016 model year vehicles.
In contrast to the calendar year analysis presented in Table III.H.10-1
through Table III.H.10-3, the model year lifetime analysis shows the
lifetime impacts of the program on each of these MY fleets over the
course of its lifetime. Full details of the inputs to this analysis can
be found in RIA Chapter 5. The societal benefits of the full life of
each of the five model years from 2012 through 2016 are shown in Tables
III.H.10-4 and III.H.10-5 at both a 3 percent and a 7 percent discount
rate, respectively. The net benefits are shown in Tables III.H.10-6 and
III.H.10-7 for both a 3 percent and a 7 percent discount rate. Note
that the quantified annual benefits shown in Table III.H.10-4 and Table
III.H.10-5 include fuel savings as a positive benefit. As such, the
quantified annual costs as shown in Table III.H.10-6 and Table
III.H.10-7 do not include fuel savings since those are included as
benefits. Also note that each of the Tables III.H.10-4 through Table
III.H.10-7 include the benefits of reduced CO2 emissions--
and consequently the total benefits--for each of four SCC values
considered by EPA. As noted above, the benefit estimates from
CO2 reductions are ``very likely,'' according to the IPCC
Fourth Assessment Report, underestimates because, in part, models used
to calculate SCC values do not include information about impacts that
have not been quantified.

Table III.H.10-4--Estimated Societal Benefits Associated with the Lifetimes of 2012-2016 Model Year Vehicles
[Millions of 2007 dollars; 3% discount rate]
----------------------------------------------------------------------------------------------------------------
Monetized values (millions) 2012MY 2013MY 2014MY 2015MY 2016MY Sum
----------------------------------------------------------------------------------------------------------------
Cost of Noise, Accident, -$1,100 -$1,600 -$2,100 -$2,900 -$3,900 -$11,600
Congestion ($).............
Pretax Fuel Savings ($)..... 16,100 23,900 32,200 46,000 63,500 181,800
Energy Security (price 900 1,400 1,800 2,500 3,500 10,100
shock) ($) a...............
Value of Reduced Refueling 1,100 1,600 2,100 3,000 4,000 11,900
time ($)...................
Value of Additional Driving 2,400 3,400 4,400 6,000 7,900 24,000
($)........................
Value of PM2.5-related 700 900 1,300 1,800 2,400 7,000
Health Impacts ($) b c d...
----------------------------------------------------------------------------------------------------------------
Reduced CO2 Emissions at each assumed SCC value e f g
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 400 500 700 1,000 1,300 3,800
Avg SCC at 3%............... 1,700 2,400 3,100 4,400 5,900 17,000
Avg SCC at 2.5%............. 2,700 3,900 5,200 7,200 9,700 29,000
95th percentile SCC at 3%... 5,100 7,300 9,600 13,000 18,000 53,000
----------------------------------------------------------------------------------------------------------------
Total Benefits at each assumed SCC value e f g
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 20,500 30,100 40,400 57,400 78,700 227,000
Avg SCC at 3%............... 21,800 32,000 42,800 60,800 83,300 240,200
Avg SCC at 2.5%............. 22,800 33,500 44,900 63,600 87,100 252,200
95th percentile SCC at 3%... 25,200 36,900 49,300 69,400 95,400 276,200
----------------------------------------------------------------------------------------------------------------
a Note that, due to a calculation error in the proposal, the energy security impacts for the model year analysis
were roughly half what they should have been.
b Note that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts. Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only
human health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental
benefits would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality
modeling. However, EPA was unable to conduct a full-scale air quality modeling analysis associated with the
vehicle model year lifetimes for the final rule.
c The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table are based on an
estimate of premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton
estimates were based on the Six Cities study (Laden et al., 2006), the values would be approximately 145%
(nearly two-and-a-half times) larger.

[[Page 25538]]

d The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table assume a 3% discount
rate in the valuation of premature mortality to account for a twenty-year segmented cessation lag. If a 7%
discount rate had been used, the values would be approximately 9% lower.
e Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used
to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
f Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
g Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC
estimates range as follows: For Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $36-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also presents these SCC estimates.

Table III.H.10-5--Estimated Societal Benefits Associated with the Lifetimes of 2012-2016 Model Year Vehicles
[Millions of 2007 dollars; 7% discount rate]
----------------------------------------------------------------------------------------------------------------
Monetized values (millions) 2012MY 2013MY 2014MY 2015MY 2016MY Sum
----------------------------------------------------------------------------------------------------------------
Cost of Noise, Accident, -$900 -$1,200 -$1,600 -$2,300 -$3,100 -$9,200
Congestion ($).............
Pretax Fuel Savings ($)..... 12,500 18,600 25,100 36,000 49,600 141,900
Energy Security (price 800 1,100 1,400 2,000 2,700 8,000
shock) ($) a...............
Value of Reduced Refueling 900 1,300 1,700 2,400 3,200 9,400
time ($)...................
Value of Additional Driving 1,900 2,700 3,500 4,700 6,200 19,000
($)........................
Value of PM2.5-related 500 800 1,000 1,400 1,900 5,600
Health Impacts ($) b c d...
----------------------------------------------------------------------------------------------------------------
Reduced CO2 Emissions at each assumed SCC value e f g
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 400 500 700 1,000 1,300 3,800
Avg SCC at 3%............... 1,700 2,400 3,100 4,400 5,900 17,000
Avg SCC at 2.5%............. 2,700 3,900 5,200 7,200 9,700 29,000
95th percentile SCC at 3%... 5,100 7,300 9,600 13,000 18,000 53,000
----------------------------------------------------------------------------------------------------------------
Total Benefits at each assumed SCC value e f g
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 16,100 23,800 31,800 45,200 61,800 178,500
Avg SCC at 3%............... 17,400 25,700 34,200 48,600 66,400 191,700
Avg SCC at 2.5%............. 18,400 27,200 36,300 51,400 70,200 203,700
95th percentile SCC at 3%... 20,800 30,600 40,700 57,200 78,500 227,700
----------------------------------------------------------------------------------------------------------------
a Note that, due to a calculation error in the proposal, the energy security impacts for the model year analysis
were roughly half what they should have been.
b Note that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts. Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only
human health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental
benefits would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality
modeling. However, EPA was unable to conduct a full-scale air quality modeling analysis associated with the
vehicle model year lifetimes for the final rule.
c The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table are based on an
estimate of premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton
estimates were based on the Six Cities study (Laden et al., 2006), the values would be approximately 145%
(nearly two-and-a-half times) larger.
d The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table assume a 3% discount
rate in the valuation of premature mortality to account for a twenty-year segmented cessation lag. If a 7%
discount rate had been used, the values would be approximately 9% lower.
e Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used
to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
f Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
g Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC
estimates range as follows: For Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $36-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also presents these SCC estimates.

Table III.H.10-6--Quantified Net Benefits Associated with the Lifetimes of 2012-2016 Model Year Vehicles
[Millions of 2007 dollars; 3% discount rate]
----------------------------------------------------------------------------------------------------------------
Monetized Values (millions) 2012MY 2013MY 2014MY 2015MY 2016MY Sum
----------------------------------------------------------------------------------------------------------------
Quantified Annual Costs $4,900 $8,000 $10,300 $12,700 $15,600 $51,500
(excluding fuel savings) a.
----------------------------------------------------------------------------------------------------------------
Quantified Annual Benefits at each assumed SCC value b c d
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 20,500 30,100 40,400 57,400 78,700 227,000
Avg SCC at 3%............... 21,800 32,000 42,800 60,800 83,300 240,200

[[Page 25539]]

Avg SCC at 2.5%............. 22,800 33,500 44,900 63,600 87,100 252,200
95th percentile SCC at 3%... 25,200 36,900 49,300 69,400 95,400 276,200
----------------------------------------------------------------------------------------------------------------
Quantified Net Benefits at each assumed SCC value b c d
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 15,600 22,100 30,100 44,700 63,100 175,500
Avg SCC at 3%............... 16,900 24,000 32,500 48,100 67,700 188,700
Avg SCC at 2.5%............. 17,900 25,500 34,600 50,900 71,500 200,700
95th percentile SCC at 3%... 20,300 28,900 39,000 56,700 79,800 224,700
----------------------------------------------------------------------------------------------------------------
a Quantified annual costs as shown here are the increased costs for new vehicles in each given model year. Since
those costs are assumed to occur in the given model year (i.e., not over a several year time span), the
discount rate does not affect the costs.
b Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used
to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
c Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
d Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC
estimates range as follows: For Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $36-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also presents these SCC estimates.

Table III.H.10-7--Quantified Net Benefits Associated With the Lifetimes of 2012-2016 Model Year Vehicles
[Millions of 2007 dollars; 7% discount rate]
----------------------------------------------------------------------------------------------------------------
Monetized values (millions) 2012MY 2013MY 2014MY 2015MY 2016MY Sum
----------------------------------------------------------------------------------------------------------------
Quantified Annual Costs $4,900 $8,000 $10,300 $12,700 $15,600 $51,500
(excluding fuel savings)
\a\........................
----------------------------------------------------------------------------------------------------------------
Quantified Annual Benefits at each assumed SCC value b c d
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 16,100 23,800 31,800 45,200 61,800 178,500
Avg SCC at 3%............... 17,400 25,700 34,200 48,600 66,400 191,700
Avg SCC at 2.5%............. 18,400 27,200 36,300 51,400 70,200 203,700
95th percentile SCC at 3%... 20,800 30,600 40,700 57,200 78,500 227,700
----------------------------------------------------------------------------------------------------------------
Quantified Net Benefits at each assumed SCC value b c d
----------------------------------------------------------------------------------------------------------------
Avg SCC at 5%............... 11,200 15,800 21,500 32,500 46,200 127,000
Avg SCC at 3%............... 12,500 17,700 23,900 35,900 50,800 140,200
Avg SCC at 2.5%............. 13,500 19,200 26,000 38,700 54,600 152,200
95th percentile SCC at 3%... 15,900 22,600 30,400 44,500 62,900 176,200
----------------------------------------------------------------------------------------------------------------
a Quantified annual costs as shown here are the increased costs for new vehicles in each given model year. Since
those costs are assumed to occur in the given model year (i.e., not over a several year time span), the
discount rate does not affect the costs.
b Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used
to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
c Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected
under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs will
contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the difference
between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop methods to
value non-CO2 emissions in future analyses.
d Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC
estimates range as follows: For Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $36-$65; and for 95th percentile SCC at 3%: $65-$136. Section III.H.6 also presents these SCC estimates.

I. Statutory and Executive Order Reviews

1. Executive Order 12866: Regulatory Planning and Review
Under section 3(f)(1) of Executive Order (EO) 12866 (58 FR 51735,
October 4, 1993), this action is an ``economically significant
regulatory action'' because it is likely to have an annual effect on
the economy of $100 million or more. Accordingly, EPA submitted this
action to the Office of Management and Budget (OMB) for review under EO
12866 and any changes made in response to OMB recommendations have been
documented in the docket for this action.
In addition, EPA prepared an analysis of the potential costs and
benefits associated with this action. This analysis is contained in the
Final Regulatory Impact Analysis, which is available in the docket for
this rulemaking and at the docket internet address listed under
ADDRESSES above.
2. Paperwork Reduction Act
The information collection requirements in this final rule have
been

[[Page 25540]]

submitted for approval to the Office of Management and Budget (OMB)
under the Paperwork Reduction Act, 44 U.S.C. 3501 et seq., and has been
assigned OMB control number 0783.57. The information collection
requirements are not enforceable until OMB approves them.
The Agency is finalizing requirements for manufacturers to submit
information to ensure compliance with the provisions in this rule. This
includes a variety of requirements for vehicle manufacturers. Section
208(a) of the Clean Air Act requires that vehicle manufacturers provide
information the Administrator may reasonably require to determine
compliance with the regulations; submission of the information is
therefore mandatory. We will consider confidential all information
meeting the requirements of section 208(c) of the Clean Air Act.
As shown in Table III.I.2-1, the total annual burden associated
with this rule is about 39,900 hours and $5 million, based on a
projection of 33 respondents. The estimated burden for vehicle
manufacturers is a total estimate for new reporting requirements.
Burden means the total time, effort, or financial resources expended by
persons to generate, maintain, retain, or disclose or provide
information to or for a Federal agency. This includes the time needed
to review instructions; develop, acquire, install, and utilize
technology and systems for the purposes of collecting, validating, and
verifying information, processing and maintaining information, and
disclosing and providing information; adjust the existing ways to
comply with any previously applicable instructions and requirements;
train personnel to be able to respond to a collection of information;
search data sources; complete and review the collection of information;
and transmit or otherwise disclose the information.

Table III.I.2-1--Estimated Burden for Reporting and Recordkeeping
Requirements
------------------------------------------------------------------------
Annual burden
Number of respondents hours Annual costs
------------------------------------------------------------------------
33................................ 39,940 $5,001,000
------------------------------------------------------------------------

An agency may not conduct or sponsor, and a person is not required
to respond to, a collection of information unless it displays a
currently valid OMB control number. The OMB control numbers for EPA's
regulations in 40 CFR are listed in 40 CFR part 9. In addition, EPA is
amending the table in 40 CFR part 9 of currently approved OMB control
numbers for various regulations to list the regulatory citations for
the information requirements contained in this final rule.
3. Regulatory Flexibility Act
a. Overview
The Regulatory Flexibility Act (RFA) generally requires an agency
to prepare a regulatory flexibility analysis of any rule subject to
notice and comment rulemaking requirements under the Administrative
Procedure Act or any other statute unless the agency certifies that the
rule will not have a significant economic impact on a substantial
number of small entities directly subject to the rule. Small entities
include small businesses, small organizations, and small governmental
jurisdictions.
For purposes of assessing the impacts of this rule on small
entities, small entity is defined as: (1) A small business as defined
by the Small Business Administration's (SBA) regulations at 13 CFR
121.201 (see table below); (2) a small governmental jurisdiction that
is a government of a city, county, town, school district or special
district with a population of less than 50,000; and (3) a small
organization that is any not-for-profit enterprise which is
independently owned and operated and is not dominant in its field.
Table III.I.3-1 provides an overview of the primary SBA small
business categories included in the light-duty vehicle sector:

Table III.I.3-1--Primary SBA Small Business Categories in the Light-Duty Vehicle Sector
----------------------------------------------------------------------------------------------------------------
Defined as small entity by
Industry \a\ SBA if less than or equal to: NAICS codes \b\
----------------------------------------------------------------------------------------------------------------
Light-duty vehicles:
--Vehicle manufacturers (including 1,000 employees.............. 336111
small volume manufacturers).
--Independent commercial importers $7 million annual sales...... 811111, 811112, 811198
$23 million annual sales..... 441120
100 employees................ 423110, 424990
--Alternative fuel vehicle 50 employees................. 336312, 336322, 336399
converters.
750 employees................ 335312
1,000 employees.............. 454312, 485310, 811198
$7 million annual sales.
----------------------------------------------------------------------------------------------------------------
Notes:
\a\ Light-duty vehicle entities that qualify as small businesses would not be subject to this rule. We are
exempting small vehicle entities, and we intend to address these entities in a future rule.
\b\ North American Industrial Classification System.

b. Summary of Potentially Affected Small Entities
EPA has not conducted a Regulatory Flexibility Analysis or a SBREFA
SBAR Panel for the rule because we are certifying that the rule would
not have a significant economic impact on a substantial number of small
entities directly subject to the rule. As proposed, EPA is exempting
manufacturers meeting SBA's business size criteria for small business
as provided in 13 CFR 121.201, due to the short lead time to develop
this rule, the extremely small emissions contribution of these
entities, and the potential need to develop a program that would be
structured

[[Page 25541]]

differently for them (which would require more time). EPA would instead
consider appropriate GHG standards for these entities as part of a
future regulatory action. This includes U.S. and foreign small entities
in three distinct categories of businesses for light-duty vehicles:
Small volume manufacturers (SVMs), independent commercial importers
(ICIs), and alternative fuel vehicle converters. EPA has identified a
total of about 47 vehicle businesses; about 13 entities (or 28 percent)
fit the Small Business Administration (SBA) criteria of a small
business. There are about 2 SVMs, 8 ICIs, and 3 alternative fuel
vehicle converters in the light-duty vehicle market which are small
businesses (no major vehicle manufacturers meet the small-entity
criteria as defined by SBA). EPA estimates that these small entities
comprise about 0.03 percent of the total light-duty vehicle sales in
the U.S., and therefore the exemption will have a negligible impact on
the GHG emissions reductions from the standards.
To ensure that EPA is aware of which companies would be exempt, EPA
proposed to require that such entities submit a declaration to EPA
containing a detailed written description of how that manufacturer
qualifies as a small entity under the provisions of 13 CFR 121.201. EPA
has reconsidered the need for this additional submission under the
regulations and is deleting it as not necessary. We already have
information on the limited number of small entities that we expect
would receive the benefits of the exemption, and do not need the
proposed regulatory requirement to be able to effectively implement
this exemption for those parties who in fact meet its terms. Small
entities are currently covered by a number of EPA motor vehicle
emission regulations, and they routinely submit information and data on
an annual basis as part of their compliance responsibilities. Based on
this, EPA is certifying that the rule would not have a significant
economic impact on a substantial number of small entities.
c. Conclusions
I therefore certify that this rule will not have a significant
economic impact on a substantial number of small entities. However, EPA
recognizes that some small entities continue to be concerned about the
potential impacts of the statutory imposition of PSD requirements that
may occur given the various EPA rulemakings currently under
consideration concerning greenhouse gas emissions. As explained in the
preamble for the proposed PSD tailoring rule (74 FR 55292, Oct. 27,
2009), EPA used the discretion afforded to it under section 609(c) of
the RFA to consult with OMB and SBA, with input from outreach to small
entities, regarding the potential impacts of PSD regulatory
requirements that might occur as EPA considers regulations of GHGs.
Concerns about the potential impacts of statutorily imposed PSD
requirements on small entities were the subject of deliberations in
that consultation and outreach. EPA has compiled a summary of that
consultation and outreach, which is available in the docket for the
Tailoring Rule (EPA-HQ-OAR-2009-0517).
4. Unfunded Mandates Reform Act
Title II of the Unfunded Mandates Reform Act of 1995 (UMRA), 2
U.S.C. 1531-1538, requires Federal agencies, unless otherwise
prohibited by law, to assess the effects of their regulatory actions on
State, local, and tribal governments and the private sector. Under
section 202 of the UMRA, EPA generally must prepare a written
statement, including a cost-benefit analysis, for proposed and final
rules with ``Federal mandates'' that may result in expenditures to
State, local, and tribal governments, in the aggregate, or to the
private sector, of $100 million or more in any one year.
This rule is not subject to the requirements of section 203 of UMRA
because it contains no regulatory requirements that might significantly
or uniquely affect small governments. This rule contains no Federal
mandates (under the regulatory provisions of Title II of the UMRA) for
State, local, or tribal governments. The rule imposes no enforceable
duty on any State, local or tribal governments. EPA has determined that
this rule contains no regulatory requirements that might significantly
or uniquely affect small governments. EPA has determined that this rule
contains a Federal mandate that may result in expenditures of $100
million or more for the private sector in any one year. EPA believes
that the action represents the least costly, most cost-effective
approach to achieve the statutory requirements of the rule. The costs
and benefits associated with the rule are discussed above and in the
Final Regulatory Impact Analysis, as required by the UMRA.
5. Executive Order 13132 (Federalism)
This action does not have federalism implications. It will not have
substantial direct effects on the States, on the relationship between
the national government and the States, or on the distribution of power
and responsibilities among the various levels of government, as
specified in Executive Order 13132. This rulemaking applies to
manufacturers of motor vehicles and not to State or local governments.
Thus, Executive Order 13132 does not apply to this action. Although
section 6 of Executive Order 13132 does not apply to this action, EPA
did consult with representatives of State governments in developing
this action.
In the spirit of Executive Order 13132, and consistent with EPA
policy to promote communications between EPA and State and local
governments, EPA specifically solicited comment on the proposed action
from State and local officials. Many State and local governments
submitted public comments on the rule, the majority of which were
supportive of the EPA's greenhouse gas program. However, these entities
did not provide comments indicating there would be a substantial direct
effect on State or local governments resulting from this rule.
6. Executive Order 13175 (Consultation and Coordination With Indian
Tribal Governments)
This action does not have tribal implications, as specified in
Executive Order 13175 (65 FR 67249, November 9, 2000). This rule will
be implemented at the Federal level and impose compliance costs only on
vehicle manufacturers. Tribal governments will be affected only to the
extent they purchase and use regulated vehicles. Thus, Executive Order
13175 does not apply to this action.
7. Executive Order 13045: ``Protection of Children From Environmental
Health Risks and Safety Risks''
This action is subject to EO 13045 (62 FR 19885, April 23, 1997)
because it is an economically significant regulatory action as defined
by EO 12866, and EPA believes that the environmental health or safety
risk addressed by this action may have a disproportionate effect on
children. A synthesis of the science and research regarding how climate
change may affect children and other vulnerable subpopulations is
contained in the Technical Support Document for Endangerment or Cause
or Contribute Findings for Greenhouse Gases under Section 202(a) of the
Clean Air Act, which can be found in the public docket for this
rule.\502\ A summary of the analysis is presented below.
---------------------------------------------------------------------------

\502\ U.S. EPA. (2009). Technical Support Document for
Endangerment or Cause or Contribute Findings for Greenhouse Gases
under Section 202(a) of the Clean Air Act. Washington, DC: U.S. EPA.
Docket EPA-HQ-OAR-2009-0472-11292.
---------------------------------------------------------------------------

With respect to GHG emissions, the effects of climate change
observed to

[[Page 25542]]

date and projected to occur in the future include the increased
likelihood of more frequent and intense heat waves. Specifically, EPA's
analysis of the scientific assessment literature has determined that
severe heat waves are projected to intensify in magnitude, frequency,
and duration over the portions of the U.S. where these events already
occur, with potential increases in mortality and morbidity, especially
among the young, elderly, and frail. EPA has estimated reductions in
projected global mean surface temperatures as a result of reductions in
GHG emissions associated with the standards finalized in this action
(Section III.F). Children may receive benefits from reductions in GHG
emissions because they are included in the segment of the population
that is most vulnerable to extreme temperatures.
For non-GHG pollutants, EPA has determined that climate change is
expected to increase regional ozone pollution, with associated risks in
respiratory infection, aggravation of asthma, and premature death. The
directional effect of climate change on ambient PM levels remains
uncertain. However, disturbances such as wildfires are increasing in
the U.S. and are likely to intensify in a warmer future with drier
soils and longer growing seasons. PM emissions from forest fires can
contribute to acute and chronic illnesses of the respiratory system,
particularly in children, including pneumonia, upper respiratory
diseases, asthma and chronic obstructive pulmonary diseases.
8. Executive Order 13211 (Energy Effects)
This rule is not a ``significant energy action'' as defined in
Executive Order 13211, ``Actions Concerning Regulations That
Significantly Affect Energy Supply, Distribution, or Use'' (66 FR 28355
(May 22, 2001)) because it is not likely to have a significant adverse
effect on the supply, distribution, or use of energy. In fact, this
rule has a positive effect on energy supply and use. Because the GHG
emission standards finalized today result in significant fuel savings,
this rule encourages more efficient use of fuels. Therefore, we have
concluded that this rule is not likely to have any adverse energy
effects. Our energy effects analysis is described above in Section
III.H.
9. National Technology Transfer Advancement Act
Section 12(d) of the National Technology Transfer and Advancement
Act of 1995 (``NTTAA''), Public Law 104-113, 12(d) (15 U.S.C. 272 note)
directs EPA to use voluntary consensus standards in its regulatory
activities unless to do so would be inconsistent with applicable law or
otherwise impractical. Voluntary consensus standards are technical
standards (e.g., materials specifications, test methods, sampling
procedures, and business practices) that are developed or adopted by
voluntary consensus standards bodies. NTTAA directs EPA to provide
Congress, through OMB, explanations when the Agency decides not to use
available and applicable voluntary consensus standards.
The rulemaking involves technical standards. Therefore, the Agency
conducted a search to identify potentially applicable voluntary
consensus standards. For CO2, N2O, and
CH4 emissions, we identified no such standards, and none
were brought to our attention in comments. Therefore, EPA is collecting
data over the same test cycles that are used for the CAFE program
following standardized test methods and sampling procedures. This will
minimize the amount of testing done by manufacturers, since
manufacturers are already required to run these tests. For A/C system
leakage improvement credits, EPA identified a Society of Automotive
Engineers (SAE) methodology and EPA's approach is based closely on this
SAE methodology. For the A/C system efficiency improvement credits,
including the new idle test, EPA generally uses standardized test
methods and sampling procedures. However, EPA knows of no consensus
standard available for an A/C idle test to measure system efficiency
improvements.
10. Executive Order 12898: Federal Actions To Address Environmental
Justice in Minority Populations and Low-Income Populations
Executive Order (EO) 12898 (59 FR 7629 (Feb. 16, 1994)) establishes
Federal executive policy on environmental justice. Its main provision
directs Federal agencies, to the greatest extent practicable and
permitted by law, to make environmental justice part of their mission
by identifying and addressing, as appropriate, disproportionately high
and adverse human health or environmental effects of their programs,
policies, and activities on minority populations and low-income
populations in the United States.
With respect to GHG emissions, EPA has determined that this final
rule will not have disproportionately high and adverse human health or
environmental effects on minority or low-income populations because it
increases the level of environmental protection for all affected
populations without having any disproportionately high and adverse
human health or environmental effects on any population, including any
minority or low-income population. The reductions in CO2 and
other GHGs associated with the standards will affect climate change
projections, and EPA has estimated reductions in projected global mean
surface temperatures (Section III.F.3). Within communities experiencing
climate change, certain parts of the population may be especially
vulnerable; these include the poor, the elderly, those already in poor
health, the disabled, those living alone, and/or indigenous populations
dependent on one or a few resources.\503\ In addition, the U.S. Climate
Change Science Program \504\ stated as one of its conclusions: ``The
United States is certainly capable of adapting to the collective
impacts of climate change. However, there will still be certain
individuals and locations where the adaptive capacity is less and these
individuals and their communities will be disproportionally impacted by
climate change.'' Therefore, these specific sub-populations may receive
benefits from reductions in GHGs.
---------------------------------------------------------------------------

\503\ U.S. EPA. (2009). Technical Support Document for
Endangerment or Cause or Contribute Findings for Greenhouse Gases
under Section 202(a) of the Clean Air Act. Washington, DC: U.S. EPA.
Docket EPA-HQ-OAR-2009-0472-11292.
\504\ CCSP (2008) Analyses of the effects of global change on
human health and welfare and human systems. A Report by the U.S.
Climate Change Science Program and the Subcommittee on Global Change
Research. [Gamble, J.L. (ed.), K.L. Ebi, F.G. Sussman, T.J.
Wilbanks, (Authors)]. U.S. Environmental Protection Agency,
Washington, DC, USA.
---------------------------------------------------------------------------

For non-GHG co-pollutants such as ozone, PM2.5, and
toxics, EPA has concluded that it is not practicable to determine
whether there would be disproportionately high and adverse human health
or environmental effects on minority and/or low income populations from
this final rule.
11. Congressional Review Act
The Congressional Review Act, 5 U.S.C. 801 et seq., as added by the
Small Business Regulatory Enforcement Fairness Act of 1996, generally
provides that before a rule may take effect, the agency promulgating
the rule must submit a rule report, which includes a copy of the rule,
to each House of the Congress and to the Comptroller General of the
United States. EPA will submit a report containing this rule and other
required information to the U.S. Senate, the U.S. House of
Representatives, and the Comptroller General of the United States prior
to publication of the rule in the Federal Register. A Major rule cannot
take effect until 60 days after it

[[Page 25543]]

is published in the Federal Register. This action is a ``major rule''
as defined by 5 U.S.C. 804(2). This rule will be effective July 6,
2010, sixty days after date of publication in the Federal Register.

J. Statutory Provisions and Legal Authority

Statutory authority for the vehicle controls finalized today is
found in section 202(a) (which authorizes standards for emissions of
pollutants from new motor vehicles which emissions cause or contribute
to air pollution which may reasonably be anticipated to endanger public
health or welfare), 202(d), 203-209, 216, and 301 of the Clean Air Act,
42 U.S.C. 7521(a), 7521(d), 7522, 7523, 7524, 7525, 7541, 7542, 7543,
7550, and 7601.

IV. NHTSA Final Rule and Record of Decision for Passenger Car and Light
Truck CAFE Standards for MYs 2012-2016

A. Executive Overview of NHTSA Final Rule

1. Introduction
The National Highway Traffic Safety Administration (NHTSA) is
establishing Corporate Average Fuel Economy (CAFE) standards for
passenger automobiles (passenger cars) and nonpassenger automobiles
(light trucks) for model years (MY) 2012-2016. Improving vehicle fuel
economy has been long and widely recognized as one of the key ways of
achieving energy independence, energy security, and a low carbon
economy.\505\ NHTSA's CAFE standards will require passenger cars and
light trucks to meet an estimated combined average of 34.1 mpg in MY
2016. This represents an average annual increase of 4.3 percent from
the 27.6 mpg combined fuel economy level in MY 2011. NHTSA's final rule
projects total fuel savings of approximately 61 billion gallons over
the lifetimes of the vehicles sold in model years 2012-2016, with
corresponding net societal benefits of over $180 billion using a 3
percent discount rate.\506\
---------------------------------------------------------------------------

\505\ Among the reports and studies noting this point are the
following:
John Podesta, Todd Stern and Kim Batten, ``Capturing the Energy
Opportunity; Creating a Low-Carbon Economy,'' Center for American
Progress (November 2007), pp. 2, 6, 8, and 24-29, available at:
http://www.americanprogress.org/issues/2007/11/pdf/energy_
chapter.pdf (last accessed March 1, 2010).
Sarah Ladislaw, Kathryn Zyla, Jonathan Pershing, Frank
Verrastro, Jenna Goodward, David Pumphrey, and Britt Staley, ``A
Roadmap for a Secure, Low-Carbon Energy Economy; Balancing Energy
Security and Climate Change,'' World Resources Institute and Center
for Strategic and International Studies (January 2009), pp. 21-22;
available at: http://pdf.wri.org/secure_low_carbon_energy_
economy_roadmap.pdf (last accessed March 1, 2010).
Alliance to Save Energy et al., ``Reducing the Cost of
Addressing Climate Change Through Energy Efficiency (2009),
available at: http://Aceee.org/energy/climate/leg.htm (last accessed
March 1, 2010).
John DeCicco and Freda Fung, ``Global Warming on the Road; The
Climate Impact of America's Automobiles,'' Environmental Defense
(2006) pp. iv-vii; available at: http://www.edf.org/documents/5301_
Globalwarmingontheroad.pdf (last accessed March 1, 2010).
``Why is Fuel Economy Important?,'' a Web page maintained by the
Department of Energy and Environmental Protection Agency, available
at http://www.fueleconomy.gov/feg/why.shtml (last accessed March 1,
2010); Robert Socolow, Roberta Hotinski, Jeffery B. Greenblatt, and
Stephen Pacala, ``Solving The Climate Problem: Technologies
Available to Curb CO2 Emissions,'' Environment, volume
46, no. 10, 2004. pages 8-19, available at: http://
www.princeton.edu/mae/people/faculty/socolow/
ENVIRONMENTDec2004issue.pdf (last accessed March 1, 2010).
\506\ This value is based on what NHTSA refers to as ``Reference
Case'' inputs, which are based on the assumptions that NHTSA has
employed for its main analysis (as opposed to sensitivity analyses
to examine the effect of variations in the assumptions on costs and
benefits). The Reference Case inputs include fuel prices based on
the AEO 2010 Reference Case, a 3 percent discount rate, a 10 percent
rebound effect, a value for the social cost of carbon (SCC) of $21/
metric ton CO2 (in 2010, rising to $45/metric ton in
2050, at a 3 percent discount rate), etc. For a full listing of the
Reference Case input assumptions, see Section IV.C.3 below.
---------------------------------------------------------------------------

The significance accorded to improving fuel economy reflects
several factors. Conserving energy, especially reducing the nation's
dependence on petroleum, benefits the U.S. in several ways. Improving
energy efficiency has benefits for economic growth and the environment,
as well as other benefits, such as reducing pollution and improving
security of energy supply. More specifically, reducing total petroleum
use decreases our economy's vulnerability to oil price shocks. Reducing
dependence on oil imports from regions with uncertain conditions
enhances our energy security. Additionally, the emission of
CO2 from the tailpipes of cars and light trucks is one of
the largest sources of U.S. CO2 emissions.\507\ Using
vehicle technology to improve fuel economy, thereby reducing tailpipe
emissions of CO2, is one of the three main measures of
reducing those tailpipe emissions of CO2.\508\ The two other
measures for reducing the tailpipe emissions of CO2 are
switching to vehicle fuels with lower carbon content and changing
driver behavior, i.e., inducing people to drive less.
---------------------------------------------------------------------------

\507\ EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2006 (April 2008), pp. ES-4, ES-8, and 2-24. Available at
http://www.epa.gov/climatechange/emissions/usgginv_archive.html
(last accessed March 1, 2010).
\508\ Podesta et al., p. 25; Ladislaw et al. p. 21; DeCicco et
al. p. vii; ``Reduce Climate Change,'' a Web page maintained by the
Department of Energy and Environmental Protection Agency at http://
www.fueleconomy.gov/feg/climate.shtml (last accessed March 1, 2010).
---------------------------------------------------------------------------

While NHTSA has been setting fuel economy standards since the
1970s, today's action represents the first-ever joint final rule by
NHTSA with another agency, the Environmental Protection Agency. As
discussed in Section I, NHTSA's final MYs 2012-2016 CAFE standards are
part of a joint National Program. A large majority of the projected
benefits are achieved jointly with EPA's GHG rule, described in detail
above in Section III of this preamble. These final CAFE standards are
consistent with the President's National Fuel Efficiency Policy
announcement of May 19, 2009, which called for harmonized rules for all
automakers, instead of three overlapping and potentially inconsistent
requirements from DOT, EPA, and the California Air Resources Board. And
finally, the final CAFE standards and the analysis supporting them also
respond to President's Obama's January 26 memorandum regarding the
setting of CAFE standards for model years 2011 and beyond.
2. Role of Fuel Economy Improvements in Promoting Energy Independence,
Energy Security, and a Low Carbon Economy
The need to reduce energy consumption is more crucial today than it
was when EPCA was enacted in the mid-1970s. U.S. energy consumption has
been outstripping U.S. energy production at an increasing rate. Net
petroleum imports now account for approximately 57 percent of U.S.
domestic petroleum consumption, and the share of U.S. oil consumption
for transportation is approximately 71 percent.\509\ Moreover, world
crude oil production continues to be highly concentrated, exacerbating
the risks of supply disruptions and their negative effects on both the
U.S. and global economies.
---------------------------------------------------------------------------

\509\ Energy Information Administration, Petroleum Basic
Statistics, updated July 2009. Available at http://www.eia.doe.gov/
basics/quickoil.html (last accessed March 1, 2010).
---------------------------------------------------------------------------

Gasoline consumption in the U.S. has historically been relatively
insensitive to fluctuations in both price and consumer income, and
people in most parts of the country tend to view gasoline consumption
as a non-discretionary expense. Thus, when gasoline's share in consumer
expenditures rises, the public experiences fiscal distress. This fiscal
distress can, in some cases, have macroeconomic consequences for the

[[Page 25544]]

economy at large. Additionally, since U.S. oil production is only
affected by fluctuations in prices over a period of years, any changes
in petroleum consumption (as through increased fuel economy) largely
flow into changes in the quantity of imports. Since petroleum imports
account for about 2 percent of GDP, increase in oil imports can create
a discernable fiscal drag. As a consequence, measures that reduce
petroleum consumption, such as fuel economy standards, will directly
benefit the balance-of-payments account, and strengthen the domestic
economy to some degree. And finally, U.S. foreign policy has been
affected for decades by rising U.S. and world dependency of crude oil
as the basis for modern transportation systems, although fuel economy
standards have only an indirect and general impact on U.S. foreign
policy.
The benefits of a low carbon economy are manifold. The U.S.
transportation sector is a significant contributor to total U.S. and
global anthropogenic emissions of greenhouse gases. Motor vehicles are
the second largest greenhouse gas-emitting sector in the U.S., after
electricity generation, and accounted for 24 percent of total U.S.
greenhouse gas emissions in 2006. Concentrations of greenhouse gases
are at unprecedented levels compared to the recent and distant past,
which means that fuel economy improvements to reduce those emissions
are a crucial step toward addressing the risks of global climate
change. These risks are well documented in Section III of this notice.
3. The National Program
NHTSA and EPA are each announcing final rules that have the effect
of addressing the urgent and closely intertwined challenges of energy
independence and security and global warming. These final rules call
for a strong and coordinated Federal greenhouse gas and fuel economy
program for passenger cars, light-duty-trucks, and medium-duty
passenger vehicles (hereafter light-duty vehicles), referred to as the
National Program. The final rules represent a coordinated program that
can achieve substantial reductions of greenhouse gas (GHG) emissions
and improvements in fuel economy from the light-duty vehicle part of
the transportation sector, based on technology that will be
commercially available and that can be incorporated at a reasonable
cost in the rulemaking timeframe. The agencies' final rules will also
provide regulatory certainty and consistency for the automobile
industry by setting harmonized national standards. They were developed
and are designed in ways that recognize and accommodate the relatively
short amount of lead time for the model years covered by the rulemaking
and the serious current economic situation faced by this industry.
These joint standards are consistent with the President's
announcement on May 19, 2009 of a National Fuel Efficiency Policy that
will reduce greenhouse gas emissions and improve fuel economy for all
new cars and light-duty trucks sold in the United States,\510\ and with
the Notice of Upcoming Joint Rulemaking signed by DOT and EPA on that
date.\511\ This joint final rule also responds to the President's
January 26, 2009 memorandum on CAFE standards for model years 2011 and
beyond, the details of which can be found below.
---------------------------------------------------------------------------

\510\ President Obama Announces National Fuel Efficiency Policy,
The White House, May 19, 2009. Available at http://
www.whitehouse.gov/the_press_office/President-Obama-Announces-
National-Fuel-Efficiency-Policy/ (last accessed March 15, 2010).
\511\ 74 FR 24007 (May 22, 2009).
---------------------------------------------------------------------------

a. Building Blocks of the National Program
The National Program is both needed and possible because the
relationship between improving fuel economy and reducing CO2
tailpipe emissions is a very direct and close one. CO2 is
the natural by-product of the combustion of fuel in motor vehicle
engines. The more fuel efficient a vehicle is, the less fuel it burns
to travel a given distance. The less fuel it burns, the less
CO2 it emits in traveling that distance.\512\ Since the
amount of CO2 emissions is essentially constant per gallon
combusted of a given type of fuel, the amount of fuel consumption per
mile is directly related to the amount of CO2 emissions per
mile. In the real world, there is a single pool of technologies for
reducing fuel consumption and CO2 emissions. Using those
technologies in the way that minimizes fuel consumption also minimizes
CO2 emissions. While there are emission control technologies
that can capture or destroy the pollutants (e.g., carbon monoxide) that
are produced by imperfect combustion of fuel, there is at present no
such technology for CO2. In fact, the only way at present to
reduce tailpipe emissions of CO2 is by reducing fuel
consumption. The National Program thus has dual benefits: it conserves
energy by improving fuel economy, as required of NHTSA by EPCA and
EISA; in the process, it necessarily reduces tailpipe CO2
emissions consonant with EPA's purposes and responsibilities under the
Clean Air Act.
---------------------------------------------------------------------------

\512\ Panel on Policy Implications of Greenhouse Warming,
National Academy of Sciences, National Academy of Engineering,
Institute of Medicine, ``Policy Implications of Greenhouse Warming:
Mitigation, Adaptation, and the Science Base,'' National Academies
Press, 1992, at 287.
---------------------------------------------------------------------------

i. DOT's CAFE Program
In 1975, Congress enacted the Energy Policy and Conservation Act
(EPCA), mandating a regulatory program for motor vehicle fuel economy
to meet the various facets of the need to conserve energy, including
ones having energy independence and security, environmental and foreign
policy implications. EPCA allocates the responsibility for implementing
the program between NHTSA and EPA as follows:
NHTSA sets Corporate Average Fuel Economy (CAFE) standards
for passenger cars and light trucks.
Because fuel economy performance is measured during
emissions regulation testing, EPA establishes the procedures for
testing, tests vehicles, collects and analyzes manufacturers' test
data, and calculates the average fuel economy of each manufacturer's
passenger cars and light trucks. EPA determines fuel economy by
measuring the amount of CO2 emitted from the tailpipe,
rather than by attempting to measure directly the amount of fuel
consumed during a vehicle test, a difficult task to accomplish with
precision. EPA then uses the carbon content of the test fuel \513\ to
calculate the amount of fuel that had to be consumed per mile in order
to produce that amount of CO2. Finally, EPA converts that
fuel consumption figure into a miles-per-gallon figure.
---------------------------------------------------------------------------

\513\ This is the method that EPA uses to determine compliance
with NHTSA's CAFE standards.
---------------------------------------------------------------------------

Based on EPA's calculation, NHTSA enforces the CAFE
standards.
The CAFE standards and compliance testing cannot capture all of the
real world CO2 emissions, because EPCA currently requires
EPA to use the 1975 passenger car test procedures under which vehicle
air conditioners are not turned on during fuel economy testing.\514\
CAFE standards also do not address the 5-8 percent of GHG emissions
that are not CO2, i.e., nitrous oxide (N2O), and
methane (CH4) as well as emissions of hydrofluorocarbons
(HFCs) related to operation of the air conditioning system.
---------------------------------------------------------------------------

\514\ See 49 U.S.C. 32904(c).
---------------------------------------------------------------------------

NHTSA has been setting CAFE standards pursuant to EPCA since the
enactment of the statute. Fuel economy gains since 1975, due both to
the standards and to market factors, have resulted in saving billions
of barrels of oil and avoiding billions of metric tons

[[Page 25545]]

of CO2 emissions. In December 2007, Congress enacted the
Energy Independence and Securities Act (EISA), amending EPCA to
require, among other things, attribute-based standards for passenger
cars and light trucks. The most recent CAFE rulemaking action was the
issuance of standards governing model years 2011 cars and trucks.
ii. EPA's Greenhouse Gas Program
On April 2, 2007, the U.S. Supreme Court issued its opinion in
Massachusetts v. EPA,\515\ a case involving a 2003 order of the
Environmental Protection Agency (EPA) denying a petition for rulemaking
to regulate greenhouse gas emissions from motor vehicles under the
Clean Air Act.\516\ The Court ruled that greenhouse gases are
``pollutants'' under the CAA and that the Act therefore authorizes EPA
to regulate greenhouse gas emissions from motor vehicles if that agency
makes the necessary findings and determinations under section 202 of
the Act. The Court considered EPCA only briefly, stating that the two
obligations may overlap, but there is no reason to think the two
agencies cannot both administer their obligations and yet avoid
inconsistency.
---------------------------------------------------------------------------

\515\ 127 S.Ct. 1438 (2007).
\516\ 68 FR 52922 (Sept. 8, 2003).
---------------------------------------------------------------------------

EPA has been working on appropriate responses that are consistent
with the decision of the Supreme Court in Massachusetts v. EPA.\517\ As
part of those responses, in July 2008, EPA issued an Advance Notice of
Proposed Rulemaking seeking comments on the impact of greenhouse gases
on the environment and on ways to reduce greenhouse gas emissions from
motor vehicles. EPA recently also issued a final rule finding that
emissions of GHGs from new motor vehicles and motor vehicle engines
cause or contribute to air pollution that endanger public health and
welfare.\518\
---------------------------------------------------------------------------

\517\ 549 U.S. 497 (2007). For further information on
Massachusetts v. EPA see the July 30, 2008 Advance Notice of
Proposed Rulemaking, ``Regulating Greenhouse Gas Emissions under the
Clean Air Act,'' 73 FR 44354 at 44397. There is a comprehensive
discussion of the litigation's history, the Supreme Court's
findings, and subsequent actions undertaken by the EPA from 2007-
2008 in response to the Supreme Court remand.
\518\ 74 FR 66496 (Dec. 15, 2009).
---------------------------------------------------------------------------

iii. California Air Resources Board's Greenhouse Gas Program
In 2004, the California Air Resources Board approved standards for
new light-duty vehicles, which regulate the emission of not only
CO2, but also other GHGs. Since then, thirteen states and
the District of Columbia, comprising approximately 40 percent of the
light-duty vehicle market, have adopted California's standards. These
standards apply to model years 2009 through 2016 and require
CO2 emissions levels for passenger cars and some light
trucks of 323 g/mil in 2009, decreasing to 205 g/mi in 2016, and 439 g/
mi for light trucks in 2009, decreasing to 332 g/mi in 2016. In 2008,
EPA denied a request by California for a waiver of preemption under the
CAA for its GHG emissions standards. However, consistent with another
Presidential Memorandum of January 26, 2009, EPA reconsidered the prior
denial of California's request.\519\ EPA withdrew the prior denial and
granted California's request for a waiver on June 30, 2009.\520\ The
granting of the waiver permits California's emission standards to come
into effect notwithstanding the general preemption of State emission
standards for new motor vehicles that otherwise applies under the Clean
Air Act.
---------------------------------------------------------------------------

\519\ 74 FR 66495 (Dec. 15, 2009). The endangerment finding was
challenged by industry in a filing submitted December 23, 2009; a
hearing date does not appear to have been set.
\520\ 74 FR 32744 (July 8, 2009).
---------------------------------------------------------------------------

b. The President's Announcement of National Fuel Efficiency Policy (May
2009)
The issue of three separate regulatory frameworks and overlapping
requirements for reducing fuel consumption and CO2 emissions
has been a subject of much controversy and legal disputes. On May 19,
2009 President Obama announced a National Fuel Efficiency Policy aimed
at both increasing fuel economy and reducing greenhouse gas pollution
for all new cars and trucks sold in the United States, while also
providing a predictable regulatory framework for the automotive
industry. The policy seeks to set harmonized Federal standards to
regulate both fuel economy and greenhouse gas emissions while
preserving the legal authorities of the Department of Transportation,
the Environmental Protection Agency and the State of California. The
program covers model year 2012 to model year 2016 and ultimately
requires the equivalent of an average fuel economy of 35.5 mpg in 2016,
if all CO2 reduction were achieved through fuel economy
improvements. Building on the MY 2011 standard that was set in March
2009, this represents an average of 5 percent increase in average fuel
economy each year between 2012 and 2016.
In conjunction with the President's announcement, the Department of
Transportation and the Environmental Protection Agency issued on May
19, 2009, a Notice of Upcoming Joint Rulemaking to propose a strong and
coordinated fuel economy and greenhouse gas National Program for Model
Year (MY) 2012-2016 light duty vehicles. Consistent, harmonized, and
streamlined requirements under that program hold out the promise of
delivering environmental and energy benefits, cost savings, and
administrative efficiencies on a nationwide basis that might not be
available under a less coordinated approach. The National Program makes
it possible for the standards of two different Federal agencies and the
standards of California and other states to act in a unified fashion in
providing these benefits. A harmonized approach to regulating light-
duty vehicle greenhouse gas (GHG) emissions and fuel economy is
critically important given the interdependent goals of addressing
climate change and ensuring energy independence and security.
Additionally, a harmonized approach may help to mitigate the cost to
manufacturers of having to comply with multiple sets of Federal and
State standards
4. Review of CAFE Standard Setting Methodology per the President's
January 26, 2009 Memorandum on CAFE Standards for MYs 2011 and Beyond
On May 2, 2008, NHTSA published a Notice of Proposed Rulemaking
entitled Average Fuel Economy Standards, Passenger Cars and Light
Trucks; Model Years 2011-2015, 73 FR 24352. In mid-October, the agency
completed and released a final environmental impact statement in
anticipation of issuing standards for those years. Based on its
consideration of the public comments and other available information,
including information on the financial condition of the automotive
industry, the agency adjusted its analysis and the standards and
prepared a final rule for MYs 2011-2015. On November 14, the Office of
Information and Regulatory Affairs (OIRA) of the Office of Management
and Budget concluded review of the rule as consistent with the
Order.\521\ However, issuance of the final rule was held in abeyance.
On January 7, 2009, the Department of Transportation announced that the
final rule would not be issued.
---------------------------------------------------------------------------

\521\ Record of OIRA's action can be found at http://
www.reginfo.gov/public/do/eoHistReviewSearch (last accessed March 1,
2010). To find the report on the clearance of the draft final rule,
select ``Department of Transportation'' under ``Economically
Significant Reviews Completed'' and select ``2008'' under ``Select
Calendar Year.''

---------------------------------------------------------------------------

[[Page 25546]]

a. Requests in the President's Memorandum
In light of the requirement to prescribe standards for MY 2011 by
March 30, 2009 and in order to provide additional time to consider
issues concerning the analysis used to determine the appropriate level
of standards for MYs 2012 and beyond, the President issued a memorandum
on January 26, 2009, requesting the Secretary of Transportation and
Administrator of the National Highway Traffic Safety Administration
NHTSA to divide the rulemaking into two parts: (1) MY 2011 standards,
and (2) standards for MY 2012 and beyond.
i. CAFE Standards for Model Year 2011
The request that the final rule establishing CAFE standards for MY
2011 passenger cars and light trucks be prescribed by March 30, 2009
was based on several factors. One was the requirement that the final
rule regarding fuel economy standards for a given model year must be
adopted at least 18 months before the beginning of that model year (49
U.S.C. 32902(g)(2)). The other was that the beginning of MY 2011 is
considered for the purposes of CAFE standard setting to be October 1,
2010.
ii. CAFE Standards for Model Years 2012 and Beyond
The President requested that, before promulgating a final rule
concerning the model years after model year 2011, NHTSA

[C]onsider the appropriate legal factors under the EISA, the
comments filed in response to the Notice of Proposed Rulemaking, the
relevant technological and scientific considerations, and to the
extent feasible, the forthcoming report by the National Academy of
Sciences mandated under section 107 of EISA.

In addition, the President requested that NHTSA consider whether
any provisions regarding preemption are appropriate under applicable
law and policy.
b. Implementing the President's Memorandum
In keeping with the President's remarks on January 26, 2009 for new
national policies to address the closely intertwined issues of energy
independence, energy security and climate change, and for the
initiation of serious and sustained domestic and international action
to address them, NHTSA has developed CAFE standards for MY 2012 and
beyond after collecting new information, conducting a careful review of
technical and economic inputs and assumptions, and standard setting
methodology, and completing new analyses.
The goal of the review and re-evaluation was to ensure that the
approach used for MY 2012 and thereafter would produce standards that
contribute, to the maximum extent possible under EPCA/EISA, to meeting
the energy and environmental challenges and goals outlined by the
President. We have sought to craft our program with the goal of
creating the maximum incentives for innovation, providing flexibility
to the regulated parties, and meeting the goal of making substantial
and continuing reductions in the consumption of fuel. To that end, we
have made every effort to ensure that the CAFE program for MYs 2012-
2016 is based on the best scientific, technical, and economic
information available, and that such information was developed in close
coordination with other Federal agencies and our stakeholders,
including the states and the vehicle manufacturers.
We have also re-examined EPCA, as amended by EISA, to consider
whether additional opportunities exist to improve the effectiveness of
the CAFE program. For example, EPCA authorizes increasing the amount of
civil penalties for violating the CAFE standards.\522\ Further, if the
test procedures used for light trucks were revised to provide for the
operation of air conditioning during fuel economy testing, vehicle
manufacturers would have a regulatory incentive to increase the
efficiency of air conditioning systems, thereby reducing both fuel
consumption and tailpipe emissions of CO2.\523\
---------------------------------------------------------------------------

\522\ Under 49 U.S.C. 32912(c), roughly, NHTSA may raise the
penalty amount if the agency decides that doing so will increase
energy conservation substantially without having a substantial
deleterious impact on the economy, employment, or competition among
automobile manufacturers.
\523\ Under 49 U.S.C. 32904(c), EPA must use the same procedures
for passenger automobiles that the Administrator used for model year
1975 (weighted 55 percent urban cycle and 45 percent highway cycle),
or procedures that give comparable results.
---------------------------------------------------------------------------

With respect to the President's request that NHTSA consider the
issue of preemption, NHTSA is deferring further consideration of the
preemption issue. The agency believes that it is unnecessary to address
the issue further at this time because of the consistent and
coordinated Federal standards that apply nationally under the National
Program.
As requested in the President's memorandum, NHTSA reviewed comments
received on the MY 2011 rulemaking and revisited its assumptions and
methodologies for purposes of developing the proposed MY 2012-2016
standards. For more information on how the proposed CAFE standards were
developed with those comments in mind, see the NPRM and the supporting
documents.
5. Summary of the Final MY 2012-2016 CAFE Standards
NHTSA is issuing CAFE standards that are, like the standards NHTSA
promulgated in March 2009 for MY 2011, expressed as mathematical
functions depending on vehicle footprint. Footprint is one measure of
vehicle size, and is determined by multiplying the vehicle's wheelbase
by the vehicle's average track width.\524\ Under the final CAFE
standards, each light vehicle model produced for sale in the United
States has a fuel economy target. The CAFE levels that must be met by
the fleet of each manufacturer will be determined by computing the
sales-weighted harmonic average of the targets applicable to each of
the manufacturer's passenger cars and light trucks. These targets, the
mathematical form and coefficients of which are presented later in
today's notice, appear as follows when the values of the targets are
plotted versus vehicle footprint:
---------------------------------------------------------------------------

\524\ See 49 CFR 523.2 for the exact definition of
``footprint.''
---------------------------------------------------------------------------

BILLING CODE 6560-50-P

[[Page 25547]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.021

[[Page 25548]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.022

BILLING CODE 6560-50-C
Under these final footprint-based CAFE standards, the CAFE levels
required of individual manufacturers depend, as noted above, on the mix
of vehicles sold. It is important to note that NHTSA's CAFE standards
and EPA's GHG standards will both be in effect, and each will lead to
increases in average fuel economy and CO2 emissions
reductions. The two agencies' standards together comprise the National
Program, and this discussion of costs and benefits of NHTSA's CAFE
standards does not change the fact that both the CAFE and GHG
standards, jointly, are the source of the benefits and costs of the
National Program.
Based on the forecast developed for this final rule of the MYs
2012-2016 vehicle fleet, NHTSA estimates that the targets shown above
will result in the following estimated average required CAFE levels:

Table IV.A.5-1--Estimated Average Required Fuel Economy (mpg) Under Final Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars................................. 33.3 34.2 34.9 36.2 37.8
Light Trucks................................... 25.4 26.0 26.6 27.5 28.8
----------------------------------------------------------------
Combined Cars & Trucks..................... 29.7 30.5 31.3 32.6 34.1
----------------------------------------------------------------------------------------------------------------

For the reader's reference, these miles per gallon values would be
equivalent to the following gallons per 100 miles values for passenger
cars and light trucks:

[[Page 25549]]

----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars................................. 3.00 2.93 2.86 2.76 2.65
Light Trucks................................... 3.94 3.85 3.76 3.63 3.48
----------------------------------------------------------------
Combined Cars & Trucks..................... 3.36 3.28 3.19 3.07 2.93
----------------------------------------------------------------------------------------------------------------

NHTSA estimates that average achieved fuel economy levels will
correspondingly increase through MY 2016, but that manufacturers will,
on average, undercomply \525\ in some model years and overcomply \526\
in others, reaching a combined average fuel economy of 33.7 mpg in MY
2016.\527\ Table IV.A.5-1 is the estimated required fuel economy for
the final CAFE standards while Table IV.A.5-2 includes the effects of
some manufacturers' payment of CAFE fines and use of FFV credits. In
addition, Section IV.G.4 below contains an analysis of the achieved
levels (and projected fuel savings, costs, and benefits) when the use
of FFV credits is assumed.
---------------------------------------------------------------------------

\525\ In NHTSA's analysis, ``undercompliance'' is mitigated
either through use of FFV credits, use of existing or ``banked''
credits, or through fine payment. Because NHTSA cannot consider
availability of credits in setting standards, the estimated achieved
CAFE levels presented here do not account for their use. In
contrast, because NHTSA is not prohibited from considering fine
payment, the estimated achieved CAFE levels presented here include
the assumption that BMW, Daimler (i.e., Mercedes), Porsche, and,
Tata (i.e., Jaguar and Rover) will only apply technology up to the
point that it would be less expensive to pay civil penalties.
\526\ In NHTSA's analysis, ``overcompliance'' occurs through
multi-year planning: manufacturers apply some ``extra'' technology
in early model years (e.g., MY 2014) in order to carry that
technology forward and thereby facilitate compliance in later model
years (e.g., MY 2016).
\527\ Consistent with EPCA, NHTSA has not accounted for
manufacturers' ability to earn CAFE credits for selling FFVs, carry
credits forward and back between model years, and transfer credits
between the passenger car and light truck fleets.

Table IV.A.5-2--Estimated Average Achieved Fuel Economy (mpg) Under Final Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars................................. 32.8 34.4 35.3 36.3 37.2
Light Trucks................................... 25.1 26.0 27.0 27.6 28.5
----------------------------------------------------------------
Combined Cars & Trucks..................... 29.3 30.6 31.7 32.6 33.7
----------------------------------------------------------------------------------------------------------------

For the reader's reference, these miles per gallon values would be
equivalent to the following gallons per 100 miles values for passenger
cars and light trucks:

----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars................................. 3.05 2.91 2.83 2.76 2.69
Light Trucks................................... 3.99 3.84 3.71 3.62 3.50
----------------------------------------------------------------
Combined Cars & Trucks..................... 3.42 3.27 3.15 3.06 2.97
----------------------------------------------------------------------------------------------------------------

NHTSA estimates that these fuel economy increases will lead to fuel
savings totaling 61 billion gallons during the lifetimes of vehicles
sold in MYs 2012-2016 (all following tables assume Reference Case
economic inputs):

Table IV.A.5-3--Fuel Saved (Billion Gallons) Under Final Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................... 2.4 5.2 7.2 9.4 11.4 35.7
Light Trucks...................... 1.8 3.7 5.3 6.5 8.1 25.4
-----------------------------------------------------------------------------
Combined...................... 4.2 8.9 12.5 16.0 19.5 61.0
----------------------------------------------------------------------------------------------------------------

The agency also estimates that these new CAFE standards will lead
to corresponding reductions of CO2 emissions totaling 655
million metric tons (mmt) during the useful lives of vehicles sold in
MYs 2012-2016:

Table IV.A.5-4--Avoided Carbon Dioxide Emissions (mmt) Under Final Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................... 25 54 77 101 123 380
Light Trucks...................... 19 40 57 71 88 275
-----------------------------------------------------------------------------

[[Page 25550]]

Combined...................... 44 94 134 172 210 655
----------------------------------------------------------------------------------------------------------------

The agency estimates that these fuel economy increases would
produce other benefits (e.g., reduced time spent refueling), as well as
some disbenefits (e.g., increased traffic congestion) caused by
drivers' tendency to increase travel when the cost of driving declines
(as it does when fuel economy increases). The agency has estimated the
total monetary value to society of these benefits and disbenefits, and
estimates that the final standards will produce significant benefits to
society. NHTSA estimates that, in present value terms, these benefits
would total over $180 billion over the useful lives of vehicles sold
during MYs 2012-2016:

Table IV.A.5-5--Present Value of Benefits ($billion) Under Final CAFE Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................... 6.8 15.2 21.6 28.7 35.2 107.5
Light Trucks...................... 5.1 10.7 15.5 19.4 24.3 75.0
-----------------------------------------------------------------------------
Combined...................... 11.9 25.8 37.1 48 59.5 182.5
----------------------------------------------------------------------------------------------------------------

NHTSA attributes most of these benefits--about $143 billion, as
noted above--to reductions in fuel consumption, valuing fuel (for
societal purposes) at future pretax prices in the Energy Information
Administration's (EIA's) reference case forecast from Annual Energy
Outlook (AEO) 2010. The Final Regulatory Impact Analysis (FRIA)
accompanying today's final rule presents a detailed analysis of
specific benefits of the final rule.

----------------------------------------------------------------------------------------------------------------
Monetized value (discounted)
Amount ----------------------------------------------------
3% Discount rate 7% Discount rate
----------------------------------------------------------------------------------------------------------------
Fuel savings....................... 61.0 billion gallons.. $143.0 billion........ $112.0 billion.
CO2 emissions reductions \528\..... 655 mmt............... $14.5 billion......... $14.5 billion.
----------------------------------------------------------------------------------------------------------------

NHTSA estimates that the necessary increases in technology
application will involve considerable monetary outlays, totaling $52
billion in incremental outlays (i.e., beyond those attributable to the
MY 2011 standards) by new vehicle purchasers during MYs 2012-2016:
---------------------------------------------------------------------------

\528\ We note that the net present value of reduced
CO2 emissions is calculated differently than other
benefits. The same discount rate used to discount the value of
damages from future emissions (SCC at 5 percent, 3 percent, and 2.5
percent) is used to calculate the net present value of the SCC for
internal consistency. Additionally, we note that the SCC increases
over time. See Social Cost of Carbon for Regulatory Impact Analysis
Under Executive Order 12866, Interagency Working Group on Social
Cost of Carbon, United States Government, February 2010 (available
in Docket No. NHTSA-2009-0059 for more information.

Table IV.A.5-6--Incremental Technology Outlays ($b) Under Final CAFE Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016 Total
----------------------------------------------------------------------------------------------------------------
Passenger Cars.................... 4.1 5.4 6.9 8.2 9.5 34.2
Light Trucks...................... 1.8 2.5 3.7 4.3 5.4 17.6
-----------------------------------------------------------------------------
Combined...................... 5.9 7.9 10.5 12.5 14.9 51.7
----------------------------------------------------------------------------------------------------------------

Corresponding to these outlays and, to a much lesser extent, civil
penalties that some companies are expected to pay for noncompliance,
the agency estimates that the final standards would lead to increases
in average new vehicle prices, ranging from $322 per vehicle in MY 2012
to $961 per vehicle in MY 2016:

Table IV.A.5-7--Incremental Increases in Average New Vehicle Prices ($) Under Final CAFE Standards
----------------------------------------------------------------------------------------------------------------
2012 2013 2014 2015 2016
----------------------------------------------------------------------------------------------------------------
Passenger Cars................................. 505 573 690 799 907
Light Trucks................................... 322 416 621 752 961
----------------------------------------------------------------
Combined................................... 434 513 665 782 926
----------------------------------------------------------------------------------------------------------------

[[Page 25551]]

Tables IV.A.5-8 and IV.A.5-9 below present itemized costs and
benefits for a 3 percent and a 7 percent discount rate, respectively,
for the combined fleet (passenger cars and light trucks) in each model
year and for all model years combined, again assuming Reference Case
inputs (except for the variation in discount rate). Numbers in
parentheses represent negative values.
---------------------------------------------------------------------------

\529\ See supra note 528.

Table IV.A.5-8--Itemized Cost and Benefit Estimates for the Combined Vehicle Fleet, 3% Discount Rate
--------------------------------------------------------------------------------------------------------------------------------------------------------
MY 2012 MY 2013 MY 2014 MY 2015 MY 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Costs:
Technology Costs.................................... 5,903 7,890 10,512 12,539 14,904 51,748
Benefits:
Savings in Lifetime Fuel Expenditures............... 9,265 20,178 29,083 37,700 46,823 143,048
Consumer Surplus from Additional Driving............ 696 1,504 2,150 2,754 3,387 10,491
Value of Savings in Refueling Time.................. 706 1,383 1,939 2,464 2,950 9,443
Reduction in Petroleum Market Externalities......... 545 1,154 1,630 2,080 2,543 7,952
Reduction in Climate-Related Damages from Lower CO2 921 2,025 2,940 3,840 4,804 14,528
Emissions \529\....................................
--------------------------------------------------------------------------------------------------------------------------------------------------------
Reduction in Health Damage Costs From Lower Emissions of Criteria Air Pollutants
--------------------------------------------------------------------------------------------------------------------------------------------------------
CO...................................................... 0 0 0 0 0 0
VOC..................................................... 42 76 102 125 149 494
NOX..................................................... 70 104 126 146 166 612
PM...................................................... 205 434 612 776 946 2,974
SOX..................................................... 158 332 469 598 731 2,288
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dis-Benefits From Increased Driving
--------------------------------------------------------------------------------------------------------------------------------------------------------
Congestion Costs........................................ (447) (902) (1,282) (1,633) (2,000) (6,264)
Noise Costs............................................. (9) (18) (25) (32) (39) (122)
Crash Costs............................................. (217) (430) (614) (778) (950) (2,989)
-----------------------------------------------------------------------------------------------
Total Benefits...................................... 11,936 25,840 37,132 48,040 59,509 182,457
===============================================================================================
Net Benefits.................................... 6,033 17,950 26,619 35,501 44,606 130,709
--------------------------------------------------------------------------------------------------------------------------------------------------------

Table IV.A.5-9--Itemized Cost and Benefit Estimates for the Combined Vehicle Fleet, 7% Discount Rate
--------------------------------------------------------------------------------------------------------------------------------------------------------
MY 2012 MY 2013 MY 2014 MY 2015 MY 2016 Total
--------------------------------------------------------------------------------------------------------------------------------------------------------
Costs:
Technology Costs.................................... 5,903 7,890 10,512 12,539 14,904 51,748
Benefits:
Savings in Lifetime Fuel Expenditures............... 7,197 15,781 22,757 29,542 36,727 112,004
Consumer Surplus from Additional Driving............ 542 1,179 1,686 2,163 2,663 8,233
Value of Savings in Refueling Time.................. 567 1,114 1,562 1,986 2,379 7,608
Reduction in Petroleum Market Externalities......... 432 917 1,296 1,654 2,023 6,322
Reduction in Climate-Related Damages From Lower CO2 921 2,025 2,940 3,840 4,804 14,530
Emissions \530\....................................
--------------------------------------------------------------------------------------------------------------------------------------------------------
Reduction in Health Damage Costs From Lower Emissions of Criteria Air Pollutants
--------------------------------------------------------------------------------------------------------------------------------------------------------
CO...................................................... 0 0 0 0 0 0
VOC..................................................... 32 60 80 99 119 390
NOx..................................................... 53 80 98 114 131 476
PM...................................................... 154 336 480 611 748 2,329
SOx..................................................... 125 265 373 475 581 1,819
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dis-Benefits From Increased Driving
--------------------------------------------------------------------------------------------------------------------------------------------------------
Congestion Costs........................................ (355) (719) (1,021) (1,302) (1,595) (4,992)
Noise Costs............................................. (7) (14) (20) (26) (31) (98)
Crash Costs............................................. (173) (342) (488) (619) (756) (2,378)
-----------------------------------------------------------------------------------------------

[[Page 25552]]

Total Benefits...................................... 9,488 20,682 29,743 38,537 47,793 146,243
===============================================================================================
Net Benefits.................................... 3,586 12,792 19,231 25,998 32,890 94,497
--------------------------------------------------------------------------------------------------------------------------------------------------------

Neither EPCA nor EISA requires that NHTSA conduct a cost-benefit
analysis in determining average fuel economy standards, but too,
neither precludes its use.\531\ EPCA does require that NHTSA consider
economic practicability among other factors, and NHTSA has concluded,
as discussed elsewhere herein, that the standards it promulgates today
are economically practicable. Further validating and supporting its
conclusion that the standards it promulgates today are reasonable, a
comparison of the standards' costs and benefits shows that the
standards' estimated benefits far outweigh its estimated costs. Based
on the figures reported above, NHTSA estimates that the total benefits
of today's final standards would be more than three times the magnitude
of the corresponding costs, such that the final standards would produce
net benefits of over $130 billion over the useful lives of vehicles
sold during MYs 2012-2016.
---------------------------------------------------------------------------

\530\ See supra note 529.
\531\ Center for Biological Diversity v. NHTSA, 508 F.3d 508
(9th Cir. 2007) (rejecting argument that EPCA precludes the use of a
marginal cost-benefit analysis that attempted to weigh all of the
social benefits (i.e., externalities as well as sdirect benefits to
consumers) of improved fuel savings in determining the stringency of
the CAFE standards). See also Entergy Corp. v. Riverkeeper, Inc.,
129 S.Ct. 1498, 1508 (2009) (``[U]nder Chevron, that an agency is
not required to [conduct a cost-benefit analysis] does not mean that
an agency is not permitted to do so.'')
---------------------------------------------------------------------------

B. Background

1. Chronology of Events Since the National Academy of Sciences Called
for Reforming and Increasing CAFE Standards
a. National Academy of Sciences Issues Report on Future of CAFE Program
(February 2002)
i. Significantly Increasing CAFE Standards Without Making Them
Attribute-Based Would Adversely Affect Safety
In the 2002 congressionally-mandated report entitled
``Effectiveness and Impact of Corporate Average Fuel Economy (CAFE)
Standards,'' \532\ a majority of the committee of the National Academy
of Sciences (NAS) (``2002 NAS Report'') concluded that the then-
existing form of passenger car and light truck CAFE standards permitted
vehicle manufacturers to comply in part by downweighting and even
downsizing their vehicles and that these actions had led to additional
fatalities. The committee explained that this safety problem arose
because, at that time, the CAFE standards were not attribute-based and
thus subjected all passenger cars to the same fuel economy target and
all light trucks to the same target, regardless of their weight, size,
or load-carrying capacity.\533\ The committee said that this experience
suggests that consideration should be given to developing a new system
of fuel economy targets that reflects differences in such vehicle
attributes. Without a thoughtful restructuring of the program, there
would be trade-offs that must be made if CAFE standards were increased
by any significant amount.\534\
---------------------------------------------------------------------------

\532\ National Research Council, ``Effectiveness and Impact of
Corporate Average Fuel Economy (CAFE) Standards,'' National Academy
Press, Washington, DC (2002). Available at http://www.nap.edu/
openbook.php?isbn=0309076013 (last accessed March 1, 2010). The
conference committee report for the Department of Transportation and
Related Agencies Appropriations Act for FY 2001 (Pub. L. 106-346)
directed NHTSA to fund a study by NAS to evaluate the effectiveness
and impacts of CAFE standards (H. Rep. No. 106-940, p. 117-118). In
response to the direction from Congress, NAS published this lengthy
report.
\533\ NHTSA formerly used this approach for CAFE standards. EISA
prohibits its use after MY 2010.
\534\ NAS, p. 9. As discussed at length in prior CAFE rules, two
members of the NAS Committee dissented from the majority opinion
that there would be safety impacts to downweighting under a flat-
standard system.
---------------------------------------------------------------------------

In response to these conclusions, NHTSA considered various
attributes and ultimately issued footprint-based CAFE standards for
light trucks and sought legislative authority to issue attribute-based
CAFE standards for passenger cars before undertaking to raise the car
standards. Congress went a step further in enacting EISA, not only
authorizing the issuance of attribute-based standards, but also
mandating them.
ii. Climate Change and Other Externalities Justify Increasing the CAFE
Standards
The NAS committee said that there are two compelling concerns that
justify increasing the fuel economy standards, both relating to
externalities. The first and most important concern, it argued, is the
accumulation in the atmosphere of greenhouse gases, principally carbon
dioxide.\535\
---------------------------------------------------------------------------

\535\ NAS, pp. 2, 13, and 83.
---------------------------------------------------------------------------

A second concern is that petroleum imports have been steadily
rising because of the nation's increasing demand for gasoline without a
corresponding increase in domestic supply. The high cost of oil imports
poses two risks: downward pressure on the strength of the dollar (which
drives up the cost of goods that Americans import) and an increase in
U.S. vulnerability to macroeconomic shocks that cost the economy
considerable real output.
To determine how much the fuel economy standards should be
increased, the committee urged that all social benefits of such
increases be considered. That is, it urged not only that the dollar
value of the saved fuel be considered, but also that the dollar value
to society of the resulting reductions in greenhouse gas emissions and
in dependence on imported oil should be calculated and considered.
iii. Reforming the CAFE Program Could Address Inequity Arising From the
CAFE Structure
The 2002 NAS report expressed concerns about increasing the
standards under the CAFE program as it was then structured. While
raising CAFE standards under the then-existing structure would reduce
fuel consumption, doing so under alternative structures ``could
accomplish the same end at lower cost, provide more flexibility to
manufacturers, or address inequities arising from the present''
structure.\536\
---------------------------------------------------------------------------

\536\ NAS, pp. 4-5 (Finding 10).
---------------------------------------------------------------------------

To address those structural problems, the report suggested various
possible reforms. The report found that the ``CAFE program might be
improved significantly by converting it to a system in which fuel
targets depend on vehicle attributes.'' \537\ The report noted further
that under an attribute-based approach, the required CAFE levels could
vary among the manufacturers based on the distribution of their product
mix. NAS

[[Page 25553]]

stated that targets could vary among passenger cars and among trucks,
based on some attribute of these vehicles such as weight, size, or
load-carrying capacity. The report explained that a particular
manufacturer's average target for passenger cars or for trucks would
depend upon the fractions of vehicles it sold with particular levels of
these attributes.\538\
---------------------------------------------------------------------------

\537\ NAS, p. 5 (Finding 12).
\538\ NAS, p. 87.
---------------------------------------------------------------------------

b. NHTSA Issues Final Rule Establishing Attribute-Based CAFE Standards
for MY 2008-2011 Light Trucks (March 2006)
The 2006 final rule reformed the structure of the CAFE program for
light trucks by introducing an attribute-based approach and using that
approach to establish higher CAFE standards for MY 2008-2011 light
trucks.\539\ Reforming the CAFE program enabled it to achieve larger
fuel savings, while enhancing safety and preventing adverse economic
consequences.
---------------------------------------------------------------------------

\539\ 71 FR 17566 (Apr. 6, 2006).
---------------------------------------------------------------------------

As noted above, fuel economy standards were restructured so that
they were based on a vehicle attribute, a measure of vehicle size
called ``footprint.'' It is the product of multiplying a vehicle's
wheelbase by its track width. A target level of fuel economy was
established for each increment in footprint (0.1 ft\2\). Trucks with
smaller footprints have higher fuel economy targets; conversely, larger
ones have lower targets. A particular manufacturer's compliance
obligation for a model year is calculated as the harmonic average of
the fuel economy targets for the manufacturer's vehicles, weighted by
the distribution of the manufacturer's production volumes among the
footprint increments. Thus, each manufacturer is required to comply
with a single overall average fuel economy level for each model year of
production.
Compared to non-attribute-based CAFE, attribute-based CAFE enhances
overall fuel savings while providing vehicle manufacturers with the
flexibility they need to respond to changing market conditions.
Attribute-based CAFE also provides a more equitable regulatory
framework by creating a level playing field for manufacturers,
regardless of whether they are full-line or limited-line manufacturers.
We were particularly encouraged that attribute-based CAFE will confer
no compliance advantage if vehicle makers choose to downsize some of
their fleet as a CAFE compliance strategy, thereby reducing the adverse
safety risks associated with the non-attribute-based CAFE program.
c. Ninth Circuit Issues Decision re Final Rule for MY 2008-2011 Light
Trucks (November 2007)
On November 15, 2007, the United States Court of Appeals for the
Ninth Circuit issued its decision in Center for Biological Diversity v.
NHTSA,\540\ the challenge to the MY 2008-11 light truck CAFE rule. The
court held that EPCA permits, but does not require, the use of a
marginal cost-benefit analysis. The court specifically emphasized
NHTSA's discretion to decide how to balance the statutory factors--as
long as that balancing does not undermine the fundamental statutory
purpose of energy conservation. Although the Court found that NHTSA had
been arbitrary and capricious in several respects, the Court did not
vacate the standards, but instead said it would remand the rule to
NHTSA to promulgate new standards consistent with its opinion ``as
expeditiously as possible and for the earliest model year
practicable.'' Under the decision, the standards established by the
April 2006 final rule would remain in effect unless and until amended
by NHTSA. In addition, it directed the agency to prepare an
Environmental Impact Statement.
---------------------------------------------------------------------------

\540\ 508 F.3d 508.
---------------------------------------------------------------------------

d. Congress Enacts Energy Security and Independence Act of 2007
(December 2007)
As noted above in Section I.B., EISA significantly changed the
provisions of EPCA governing the establishment of future CAFE
standards. These changes made it necessary for NHTSA to pause in its
efforts so that it could assess the implications of the amendments made
by EISA and then, as required, revise some aspects of the proposals it
had been developing (e.g., the model years covered and credit issues).
e. NHTSA Proposes CAFE Standards for MYs 2011-2015 (April 2008)
The agency could not set out the exact level of CAFE that each
manufacturer would have been required to meet for each model year under
the passenger car or light truck standards since the levels would
depend on information that would not be available until the end of each
of the model years, i.e., the final actual production figures for each
of those years. The agency could, however, project what the industry-
wide level of average fuel economy would have been for passenger cars
and for light trucks if each manufacturer produced its expected mix of
automobiles and just met its obligations under the proposed
``optimized'' standards for each model year.

------------------------------------------------------------------------
Passenger Light
cars mpg trucks mpg
------------------------------------------------------------------------
MY 2011....................................... 31.2 25.0
MY 2012....................................... 32.8 26.4
MY 2013....................................... 34.0 27.8
MY 2014....................................... 34.8 28.2
MY 2015....................................... 35.7 28.6
------------------------------------------------------------------------

The combined industry-wide average fuel economy (in miles per
gallon, or mpg) levels for both cars and light trucks, if each
manufacturer just met its obligations under the proposed ``optimized''
standards for each model year, would have been as follows:

------------------------------------------------------------------------
Combined
mpg
------------------------------------------------------------------------
MY 2011.................................................... 27.8
MY 2012.................................................... 29.2
MY 2013.................................................... 30.5
MY 2014.................................................... 31.0
MY 2015.................................................... 31.6
------------------------------------------------------------------------

The annual average increase during this five year period would have
been approximately 4.5 percent. Due to the uneven distribution of new
model introductions during this period and to the fact that significant
technological changes could be most readily made in conjunction with
those introductions, the annual percentage increases were greater in
the early years in this period.
f. Ninth Circuit Revises Its Decision re Final Rule for MY 2008-2011
Light Trucks (August 2008)
In response to the Government petition for rehearing, the Ninth
Circuit modified its decision by replacing its direction to prepare an
EIS with a direction to prepare either a new EA or, if necessary, an
EIS.\541\
---------------------------------------------------------------------------

\541\ See CBD v. NHTSA, 538 F.3d 1172 (9th Cir. 2008).
---------------------------------------------------------------------------

g. NHTSA Releases Final Environmental Impact Statement (October 2008)
On October 17, 2008, EPA published a notice announcing the
availability of NHTSA's final environmental impact statement (FEIS) for
the MYs 2011-2015 rulemaking.\542\ Throughout the FEIS, NHTSA relied
extensively on findings of the United Nations Intergovernmental Panel
on Climate Change (IPCC) and the U.S. Climate Change Science Program
(USCCSP). In particular, the agency relied heavily on the most recent,
thoroughly peer-reviewed, and credible assessments of global climate
change and its impact on the United States: The

[[Page 25554]]

IPCC Fourth Assessment Report Working Group I4 and II5 Reports, and
reports by the USCCSP that include Scientific Assessments of the
Effects of Global Climate Change on the United States and Synthesis and
Assessment Products.
---------------------------------------------------------------------------

\542\ 73 FR 61859 (Oct. 18, 2008).
---------------------------------------------------------------------------

In the FEIS, NHTSA compared the environmental impacts of its
preferred alternative and those of reasonable alternatives. It
considered direct, indirect, and cumulative impacts and describes these
impacts to inform the decision maker and the public of the
environmental impacts of the various alternatives.
Among other potential impacts, NHTSA analyzed the direct and
indirect impacts related to fuel and energy use, emissions, including
carbon dioxide and its effects on temperature and climate change, air
quality, natural resources, and the human environment. Specifically,
the FEIS used a climate model to estimate and report on four direct and
indirect effects of climate change, driven by alternative scenarios of
GHG emissions, including:
1. Changes in CO2 concentrations;
2. Changes in global mean surface temperature;
3. Changes in regional temperature and precipitation; and
4. Changes in sea level.
NHTSA also considered the cumulative impacts of the proposed
standards for MY 2011-2015 passenger cars and light trucks, together
with estimated impacts of NHTSA's implementation of the CAFE program
through MY 2010 and NHTSA's future CAFE rulemaking for MYs 2016-2020.
h. Department of Transportation Decides Not To Issue MY 2011-2015 Final
Rule (January 2009)
On January 7, 2009, the Department of Transportation announced that
the Bush Administration would not issue the final rule, notwithstanding
the Office of Information and Regulatory Affairs' completion of review
of the rule under Executive Order 12866, Regulatory Planning and
Review, on November 14, 2008.\543\
---------------------------------------------------------------------------

\543\ The statement can be found at http://www.dot.gov/affairs/
dot0109.htm (last accessed March 1, 2010).
---------------------------------------------------------------------------

i. The President Requests NHTSA To Issue Final Rule for MY 2011 Only
(January 2009)
As explained above, in his memorandum of January 26, 2009, the
President requested the agency to issue a final rule adopting CAFE
standards for MY 2011 only. Further, the President requested NHTSA to
establish standards for MY 2012 and later after considering the
appropriate legal factors, the comments filed in response to the May
2008 proposal, the relevant technological and scientific
considerations, and, to the extent feasible, a forthcoming report by
the National Academy of Sciences assessing automotive technologies that
can practicably be used to improve fuel economy.
j. NHTSA Issues Final Rule for MY 2011 (March 2009)
i. Standards
The final rule established footprint-based fuel economy standards
for MY 2011 passenger cars and light trucks. Each vehicle
manufacturer's required level of CAFE was based on target levels of
average fuel economy set for vehicles of different sizes and on the
distribution of that manufacturer's vehicles among those sizes. The
curves defining the performance target at each footprint reflect the
technological and economic capabilities of the industry. The target for
each footprint is the same for all manufacturers, regardless of
differences in their overall fleet mix. Compliance would be determined
by comparing a manufacturer's harmonically averaged fleet fuel economy
levels in a model year with a required fuel economy level calculated
using the manufacturer's actual production levels and the targets for
each footprint of the vehicles that it produces.
The agency analyzed seven regulatory alternatives, one of which
maximizes net benefits within the limits of available information and
was known at the time as the ``optimized standards.'' The optimized
standards were set at levels, such that, considering all of the
manufacturers together, no other alternative is estimated to produce
greater net benefits to society. Upon a considered analysis of all
information available, including all information submitted to NHTSA in
comments, the agency adopted the ``optimized standard'' alternative as
the final standards for MY 2011.\544\ By limiting the standards to
levels that can be achieved using technologies each of which are
estimated to provide benefits that at least equal its costs, the net
benefit maximization approach helped, at the time, to assure the
marketability of the manufacturers' vehicles and thus economic
practicability of the standards, for the reasons discussed extensively
in that final rule.
---------------------------------------------------------------------------

\544\ The agency notes, for NEPA purposes, that the ``optimized
standard'' alternative adopted as the final standards corresponds to
the ``Optimized Mid-2'' scenario described in Section 2.2.2 of the
FEIS.
---------------------------------------------------------------------------

The following levels were projected for what the industry-wide
level of average fuel economy will be for passenger cars and for light
trucks if each manufacturer produced its expected mix of automobiles
and just met its obligations under the ``optimized'' standards.

------------------------------------------------------------------------
Passenger Light
cars mpg trucks mpg
------------------------------------------------------------------------
MY 2011....................................... 30.2 24.1
------------------------------------------------------------------------

The combined industry-wide average fuel economy (in miles per
gallon, or mpg) levels for both cars and light trucks, if each
manufacturer just met its obligations under the ``optimized''
standards, were projected as follows:

------------------------------------------------------------------------
mpg
Combined increase
mpg over prior
year
------------------------------------------------------------------------
MY 2011....................................... 27.3 2.0
------------------------------------------------------------------------

In addition, per EISA, each manufacturer's domestic passenger fleet
is required in MY 2011 to achieve 27.5 mpg or 92 percent of the CAFE of
the industry-wide combined fleet of domestic and non-domestic passenger
cars \545\ for that model year, whichever is higher. This requirement
resulted in the following projected alternative minimum standard (not
attribute-based) for domestic passenger cars:
---------------------------------------------------------------------------

\545\ Those numbers set out several paragraphs above.

------------------------------------------------------------------------
Domestic
passenger
cars mpg
------------------------------------------------------------------------
MY 2011.................................................... 27.8
------------------------------------------------------------------------

ii. Credits
NHTSA also adopted a new part 536 on use of ``credits'' earned for
exceeding applicable CAFE standards. Part 536 implements the provisions
in EISA authorizing NHTSA to establish by regulation a credit trading
program and directing it to establish by regulation a credit transfer
program.\546\ Since its enactment, EPCA has permitted manufacturers to
earn credits for exceeding the standards and to apply those credits to
compliance obligations

[[Page 25555]]

in years other than the model year in which it was earned. EISA
extended the ``carry-forward'' period to five model years, and left the
``carry-back'' period at three model years. Under part 536, credit
holders (including, but not limited to, manufacturers) will have credit
accounts with NHTSA, and will be able to hold credits, apply them to
compliance with CAFE standards, transfer them to another ``compliance
category'' for application to compliance there, or trade them. A credit
may also be cancelled before its expiry date, if the credit holder so
chooses. Traded and transferred credits will be subject to an
``adjustment factor'' to ensure total oil savings are preserved, as
required by EISA. EISA also prohibits credits earned before MY 2011
from being transferred, so NHTSA has developed several regulatory
restrictions on trading and transferring to facilitate Congress' intent
in this regard.
---------------------------------------------------------------------------

\546\ Congress required that DOT establish a credit
``transferring'' regulation, to allow individual manufacturers to
move credits from one of their fleets to another (e.g., using a
credit earned for exceeding the light truck standard for compliance
with the domestic passenger car standard). Congress allowed DOT to
establish a credit ``trading'' regulation, so that credits may be
bought and sold between manufacturers and other parties.
---------------------------------------------------------------------------

2. Energy Policy and Conservation Act, as Amended by the Energy
Independence and Security Act
NHTSA establishes CAFE standards for passenger cars and light
trucks for each model year under EPCA, as amended by EISA. EPCA
mandates a motor vehicle fuel economy regulatory program to meet the
various facets of the need to conserve energy, including ones having
environmental and foreign policy implications. EPCA allocates the
responsibility for implementing the program between NHTSA and EPA as
follows: NHTSA sets CAFE standards for passenger cars and light trucks;
EPA establishes the procedures for testing, tests vehicles, collects
and analyzes manufacturers' data, and calculates the average fuel
economy of each manufacturer's passenger cars and light trucks; and
NHTSA enforces the standards based on EPA's calculations.
a. Standard Setting
We have summarized below the most important aspects of standard
setting under EPCA, as amended by EISA.
For each future model year, EPCA requires that NHTSA establish
standards at ``the maximum feasible average fuel economy level that it
decides the manufacturers can achieve in that model year,'' based on
the agency's consideration of four statutory factors: Technological
feasibility, economic practicability, the effect of other standards of
the Government on fuel economy, and the need of the nation to conserve
energy. EPCA does not define these terms or specify what weight to give
each concern in balancing them; thus, NHTSA defines them and determines
the appropriate weighting based on the circumstances in each CAFE
standard rulemaking.\547\
---------------------------------------------------------------------------

\547\ See Center for Biological Diversity v. NHTSA, 538 F.3d.
1172, 1195 (9th Cir. 2008) (``The EPCA clearly requires the agency
to consider these four factors, but it gives NHTSA discretion to
decide how to balance the statutory factors--as long as NHTSA's
balancing does not undermine the fundamental purpose of the EPCA:
energy conservation.'')
---------------------------------------------------------------------------

For MYs 2011-2020, EPCA further requires that separate standards
for passenger cars and for light trucks be set at levels high enough to
ensure that the CAFE of the industry-wide combined fleet of new
passenger cars and light trucks reaches at least 35 mpg not later than
MY 2020.
i. Factors That Must Be Considered in Deciding the Appropriate
Stringency of CAFE Standards
(1) Technological Feasibility
``Technological feasibility'' refers to whether a particular method
of improving fuel economy can be available for commercial application
in the model year for which a standard is being established. Thus, the
agency is not limited in determining the level of new standards to
technology that is already being commercially applied at the time of
the rulemaking. NHTSA has historically considered all types of
technologies that improve real-world fuel economy, except those whose
effects are not reflected in fuel economy testing. Principal among them
are technologies that improve air conditioner efficiency because the
air conditioners are not turned on during testing under existing test
procedures.
(2) Economic Practicability
``Economic practicability'' refers to whether a standard is one
``within the financial capability of the industry, but not so stringent
as to'' lead to ``adverse economic consequences, such as a significant
loss of jobs or the unreasonable elimination of consumer choice.''
\548\ This factor is especially important in the context of current
events, where the automobile industry is facing significantly adverse
economic conditions, as well as significant loss of jobs. In an attempt
to ensure the economic practicability of attribute-based standards,
NHTSA considers a variety of factors, including the annual rate at
which manufacturers can increase the percentage of their fleets that
employ a particular type of fuel-saving technology, and cost to
consumers. Consumer acceptability is also an element of economic
practicability, one which is particularly difficult to gauge during
times of frequently-changing fuel prices. NHTSA believes this approach
is reasonable for the MY 2012-2016 standards in view of the facts
before it at this time.
---------------------------------------------------------------------------

\548\ 67 FR 77015, 77021 (Dec. 16, 2002).
---------------------------------------------------------------------------

At the same time, the law does not preclude a CAFE standard that
poses considerable challenges to any individual manufacturer. The
Conference Report for EPCA, as enacted in 1975, makes clear, and the
case law affirms, ``a determination of maximum feasible average fuel
economy should not be keyed to the single manufacturer which might have
the most difficulty achieving a given level of average fuel economy.''
\549\ Instead, NHTSA is compelled ``to weigh the benefits to the nation
of a higher fuel economy standard against the difficulties of
individual automobile manufacturers.'' Id. The law permits CAFE
standards exceeding the projected capability of any particular
manufacturer as long as the standard is economically practicable for
the industry as a whole. Thus, while a particular CAFE standard may
pose difficulties for one manufacturer, it may also present
opportunities for another. The CAFE program is not necessarily intended
to maintain the competitive positioning of each particular company.
Rather, it is intended to enhance fuel economy of the vehicle fleet on
American roads, while protecting motor vehicle safety and being mindful
of the risk of harm to the overall United States economy.
---------------------------------------------------------------------------

\549\ CEI-I, 793 F.2d 1322, 1352 (DC Cir. 1986).
---------------------------------------------------------------------------

(3) The Effect of Other Motor Vehicle Standards of the Government on
Fuel Economy
``The effect of other motor vehicle standards of the Government on
fuel economy,'' involves an analysis of the effects of compliance with
emission,\550\ safety, noise, or damageability standards on fuel
economy capability and thus on average fuel economy. In previous CAFE
rulemakings, the agency has said that pursuant to this provision, it
considers the adverse effects of other motor vehicle standards on fuel
economy. It said so because, from the CAFE program's earliest years
\551\ until present, the effects of such compliance on fuel economy
capability over the history of the CAFE program have been negative
ones. For example, safety standards that have the effect of increasing
vehicle weight lower vehicle

[[Page 25556]]

fuel economy capability and thus decrease the level of average fuel
economy that the agency can determine to be feasible.
---------------------------------------------------------------------------

\550\ In the case of emission standards, this includes standards
adopted by the Federal government and can include standards adopted
by the States as well, since in certain circumstances the Clean Air
Act allows States to adopt and enforce State standards different
from the Federal ones.
\551\ 42 FR 63184, 63188 (Dec. 15, 1977). See also 42 FR 33534,
33537 (Jun. 30, 1977).
---------------------------------------------------------------------------

NHTSA also recognizes that in some cases the effect of other motor
vehicle standards of the Government on fuel economy may be neutral or
positive. For example, to the extent the GHG standards set by EPA and
California result in increases in fuel economy, they would do so almost
exclusively as a result of inducing manufacturers to install the same
types of technologies used by manufacturers in complying with the CAFE
standards. The primary exception would involve lower-GHG-producing air
conditioners. The agency considered EPA's standards and the
harmonization benefits of the National Program in developing its own
standards.
(4) The Need of the United States To Conserve Energy
``The need of the United States to conserve energy'' means ``the
consumer cost, national balance of payments, environmental, and foreign
policy implications of our need for large quantities of petroleum,
especially imported petroleum.'' \552\ Environmental implications
principally include reductions in emissions of criteria pollutants and
carbon dioxide. Prime examples of foreign policy implications are
energy independence and security concerns.
---------------------------------------------------------------------------

\552\ 42 FR 63184, 63188 (1977).
---------------------------------------------------------------------------

(a) Fuel Prices and the Value of Saving Fuel
Projected future fuel prices are a critical input into the
preliminary economic analysis of alternative CAFE standards, because
they determine the value of fuel savings both to new vehicle buyers and
to society. In this rule, NHTSA relies on fuel price projections from
the U.S. Energy Information Administration's (EIA) Annual Energy
Outlook (AEO) for this analysis. Federal government agencies generally
use EIA's projections in their assessments of future energy-related
policies.
(b) Petroleum Consumption and Import Externalities
U.S. consumption and imports of petroleum products impose costs on
the domestic economy that are not reflected in the market price for
crude petroleum, or in the prices paid by consumers of petroleum
products such as gasoline. These costs include (1) higher prices for
petroleum products resulting from the effect of U.S. oil import demand
on the world oil price; (2) the risk of disruptions to the U.S. economy
caused by sudden reductions in the supply of imported oil to the U.S.;
and (3) expenses for maintaining a U.S. military presence to secure
imported oil supplies from unstable regions, and for maintaining the
strategic petroleum reserve (SPR) to provide a response option should a
disruption in commercial oil supplies threaten the U.S. economy, to
allow the United States to meet part of its International Energy Agency
obligation to maintain emergency oil stocks, and to provide a national
defense fuel reserve. Higher U.S. imports of crude oil or refined
petroleum products increase the magnitude of these external economic
costs, thus increasing the true economic cost of supplying
transportation fuels above the resource costs of producing them.
Conversely, reducing U.S. imports of crude petroleum or refined fuels
or reducing fuel consumption can reduce these external costs.
(c) Air Pollutant Emissions
While reductions in domestic fuel refining and distribution that
result from lower fuel consumption will reduce U.S. emissions of
various pollutants, additional vehicle use associated with the rebound
effect \553\ from higher fuel economy will increase emissions of these
pollutants. Thus, the net effect of stricter CAFE standards on
emissions of each pollutant depends on the relative magnitudes of its
reduced emissions in fuel refining and distribution, and increases in
its emissions from vehicle use.
---------------------------------------------------------------------------

\553\ The ``rebound effect'' refers to the tendency of drivers
to drive their vehicles more as the cost of doing so goes down, as
when fuel economy improves.
---------------------------------------------------------------------------

Fuel savings from stricter CAFE standards also result in lower
emissions of CO2, the main greenhouse gas emitted as a
result of refining, distribution, and use of transportation fuels.
Lower fuel consumption reduces carbon dioxide emissions directly,
because the primary source of transportation-related CO2
emissions is fuel combustion in internal combustion engines.
NHTSA has considered environmental issues, both within the context
of EPCA and the National Environmental Policy Act, in making decisions
about the setting of standards from the earliest days of the CAFE
program. As courts of appeal have noted in three decisions stretching
over the last 20 years,\554\ NHTSA defined the ``need of the Nation to
conserve energy'' in the late 1970s as including ``the consumer cost,
national balance of payments, environmental, and foreign policy
implications of our need for large quantities of petroleum, especially
imported petroleum.'' \555\ Pursuant to that view, NHTSA declined in
the past to include diesel engines in determining the appropriate level
of standards for passenger cars and for light trucks because
particulate emissions from diesels were then both a source of concern
and unregulated.\556\ In 1988, NHTSA included climate change concepts
in its CAFE notices and prepared its first environmental assessment
addressing that subject.\557\ It cited concerns about climate change as
one of its reasons for limiting the extent of its reduction of the CAFE
standard for MY 1989 passenger cars.\558\ Since then, NHTSA has
considered the benefits of reducing tailpipe carbon dioxide emissions
in its fuel economy rulemakings pursuant to the statutory requirement
to consider the nation's need to conserve energy by reducing fuel
consumption.
---------------------------------------------------------------------------

\554\ Center for Auto Safety v. NHTSA, 793 F.2d 1322, 1325 n. 12
(DC Cir. 1986); Public Citizen v. NHTSA, 848 F.2d 256, 262-3 n. 27
(DC Cir. 1988) (noting that ``NHTSA itself has interpreted the
factors it must consider in setting CAFE standards as including
environmental effects''); and Center for Biological Diversity v.
NHTSA, 538 F.3d 1172 (9th Cir. 2007).
\555\ 42 FR 63184, 63188 (Dec. 15, 1977) (emphasis added).
\556\ For example, the final rules establishing CAFE standards
for MY 1981-84 passenger cars, 42 FR 33533, 33540-1 and 33551 (Jun.
30, 1977), and for MY 1983-85 light trucks, 45 FR 81593, 81597 (Dec.
11, 1980).
\557\ 53 FR 33080, 33096 (Aug. 29, 1988).
\558\ 53 FR 39275, 39302 (Oct. 6, 1988).
---------------------------------------------------------------------------

ii. Other Factors Considered by NHTSA
NHTSA considers the potential for adverse safety consequences when
in establishing CAFE standards. This practice is recognized approvingly
in case law.\559\ Under the universal or ``flat'' CAFE standards that
NHTSA was previously authorized to establish, manufacturers were
encouraged to respond to higher standards by building smaller, less
safe vehicles in order to ``balance out'' the larger, safer vehicles
that the public generally preferred to

[[Page 25557]]

buy, which resulted in a higher mass differential between the smallest
and the largest vehicles, with a correspondingly greater risk to
safety. Under the attribute-based standards being proposed today, that
risk is reduced because building smaller vehicles would tend to raise a
manufacturer's overall CAFE obligation, rather than only raising its
fleet average CAFE, and because all vehicles are required to continue
improving their fuel economy.
---------------------------------------------------------------------------

\559\ See, e.g., Center for Auto Safety v. NHTSA (CAS), 793 F.
2d 1322 (DC Cir. 1986) (Administrator's consideration of market
demand as component of economic practicability found to be
reasonable); Public Citizen 848 F.2d 256 (Congress established broad
guidelines in the fuel economy statute; agency's decision to set
lower standard was a reasonable accommodation of conflicting
policies). As the United States Court of Appeals pointed out in
upholding NHTSA's exercise of judgment in setting the 1987-1989
passenger car standards, ``NHTSA has always examined the safety
consequences of the CAFE standards in its overall consideration of
relevant factors since its earliest rulemaking under the CAFE
program.'' Competitive Enterprise Institute v. NHTSA (CEI I), 901
F.2d 107, 120 at n.11 (DC Cir. 1990).
---------------------------------------------------------------------------

In addition, the agency considers consumer demand in establishing
new standards and in assessing whether already established standards
remained feasible. In the 1980s, the agency relied in part on the
unexpected drop in fuel prices and the resulting unexpected failure of
consumer demand for small cars to develop in explaining the need to
reduce CAFE standards for a several year period in order to give
manufacturers time to develop alternative technology-based strategies
for improving fuel economy.
iii. Factors That NHTSA Is Statutorily Prohibited From Considering in
Setting Standards
EPCA provides that in determining the level at which it should set
CAFE standards for a particular model year, NHTSA may not consider the
ability of manufacturers to take advantage of several EPCA provisions
that facilitate compliance with the CAFE standards and thereby reduce
the costs of compliance.\560\ As noted below, manufacturers can earn
compliance credits by exceeding the CAFE standards and then use those
credits to achieve compliance in years in which their measured average
fuel economy falls below the standards. Manufacturers can also increase
their CAFE levels through MY 2019 by producing alternative fuel
vehicles. EPCA provides an incentive for producing these vehicles by
specifying that their fuel economy is to be determined using a special
calculation procedure that results in those vehicles being assigned a
high fuel economy level.
---------------------------------------------------------------------------

\560\ 49 U.S.C. 32902(h).
---------------------------------------------------------------------------

iv. Weighing and Balancing of Factors
NHTSA has broad discretion in balancing the above factors in
determining the average fuel economy level that the manufacturers can
achieve. Congress ``specifically delegated the process of setting * * *
fuel economy standards with broad guidelines concerning the factors
that the agency must consider. The breadth of those guidelines, the
absence of any statutorily prescribed formula for balancing the
factors, the fact that the relative weight to be given to the various
factors may change from rulemaking to rulemaking as the underlying
facts change, and the fact that the factors may often be conflicting
with respect to whether they militate toward higher or lower standards
give NHTSA discretion to decide what weight to give each of the
competing policies and concerns and then determine how to balance them
as long as NHTSA's balancing does not undermine the fundamental purpose
of the EPCA: Energy conservation, and as long as that balancing
reasonably accommodates `conflicting policies that were committed to
the agency's care by the statute.' ''
Thus, EPCA does not mandate that any particular number be adopted
when NHTSA determines the level of CAFE standards. Rather, any number
within a zone of reasonableness may be, in NHTSA's assessment, the
level of stringency that manufacturers can achieve. See, e.g., Hercules
Inc. v. EPA, 598 F. 2d 91, 106 (DC Cir. 1978) (``In reviewing a
numerical standard we must ask whether the agency's numbers are within
a zone of reasonableness, not whether its numbers are precisely
right'').
v. Other Requirements Related to Standard Setting
The standards for passenger cars and those for light trucks must
increase ratably each year. This statutory requirement is interpreted,
in combination with the requirement to set the standards for each model
year at the level determined to be the maximum feasible level that
manufacturers can achieve for that model year, to mean that the annual
increases should not be disproportionately large or small in relation
to each other.
The standards for passenger cars and light trucks must be based on
one or more vehicle attributes, like size or weight, that correlate
with fuel economy and must be expressed in terms of a mathematical
function. Fuel economy targets are set for individual vehicles and
increase as the attribute decreases and vice versa. For example, size-
based (i.e., size-indexed) standards assign higher fuel economy targets
to smaller (and generally, but not necessarily, lighter) vehicles and
lower ones to larger (and generally, but not necessarily, heavier)
vehicles. The fleet-wide average fuel economy that a particular
manufacturer is required to achieve depends on the size mix of its
fleet, i.e., the proportion of the fleet that is small-, medium- or
large-sized.
This approach can be used to require virtually all manufacturers to
increase significantly the fuel economy of a broad range of both
passenger cars and light trucks, i.e., the manufacturer must improve
the fuel economy of all the vehicles in its fleet. Further, this
approach can do so without creating an incentive for manufacturers to
make small vehicles smaller or large vehicles larger, with attendant
implications for safety.
b. Test Procedures for Measuring Fuel Economy
EPCA provides EPA with the responsibility for establishing CAFE
test procedures. Current test procedures measure the effects of many
fuel saving technologies. The principal exception is improvements in
air conditioning efficiency. By statutory law in the case of passenger
cars and by administrative regulation in the case of light trucks, air
conditioners are not turned on during fuel economy testing.
The fuel economy test procedures for light trucks could be amended
through rulemaking to provide for air conditioner operation during
testing and to take other steps for improving the accuracy and
representativeness of fuel economy measurements. NHTSA sought comment
in the NPRM regarding implementing such amendments beginning in MY 2017
and also on the more immediate interim alternative step of providing
CAFE program credits under the authority of 49 U.S.C. 32904(c) for
light trucks equipped with relatively efficient air conditioners for
MYs 2012-2016, but decided against finalizing either option for
purposes of this final rule, choosing to defer the matter for now.
Modernizing the passenger car test procedures, or even providing
similar credits, would not be possible under EPCA as currently written.
c. Enforcement and Compliance Flexibility
EPA is responsible for measuring automobile manufacturers' CAFE so
that NHTSA can determine compliance with the CAFE standards. When NHTSA
finds that a manufacturer is not in compliance, it notifies the
manufacturer. Surplus credits generated from the five previous years
can be used to make up the deficit. The amount of credit earned is
determined by multiplying the number of tenths of a mpg by which a
manufacturer exceeds a standard for a particular category of
automobiles by the total volume of automobiles of that category
manufactured by the manufacturer for a given model year. If there are
no (or not

[[Page 25558]]

enough) credits available, then the manufacturer can either pay the
fine, or submit a carry back plan to NHTSA. A carry back plan describes
what the manufacturer plans to do in the following three model years to
earn enough credits to make up for the deficit. NHTSA must examine and
determine whether to approve the plan.
In the event that a manufacturer does not comply with a CAFE
standard, even after the consideration of credits, EPCA provides for
the assessing of civil penalties, unless, as provided below, the
manufacturer has earned credits for exceeding a standard in an earlier
year or expects to earn credits in a later year.\561\ The Act specifies
a precise formula for determining the amount of civil penalties for
such a noncompliance. The penalty, as adjusted for inflation by law, is
$5.50 for each tenth of a mpg that a manufacturer's average fuel
economy falls short of the standard for a given model year multiplied
by the total volume of those vehicles in the affected fleet (i.e.,
import or domestic passenger car, or light truck), manufactured for
that model year. The amount of the penalty may not be reduced except
under the unusual or extreme circumstances specified in the statute.
---------------------------------------------------------------------------

\561\ EPCA does not provide authority for seeking to enjoin
violations of the CAFE standards.
---------------------------------------------------------------------------

Unlike the National Traffic and Motor Vehicle Safety Act, EPCA does
not provide for recall and remedy in the event of a noncompliance. The
presence of recall and remedy provisions \562\ in the Safety Act and
their absence in EPCA is believed to arise from the difference in the
application of the safety standards and CAFE standards. A safety
standard applies to individual vehicles; that is, each vehicle must
possess the requisite equipment or feature that must provide the
requisite type and level of performance. If a vehicle does not, it is
noncompliant. Typically, a vehicle does not entirely lack an item or
equipment or feature. Instead, the equipment or features fails to
perform adequately. Recalling the vehicle to repair or replace the
noncompliant equipment or feature can usually be readily accomplished.
---------------------------------------------------------------------------

\562\ 49 U.S.C. 30120, Remedies for defects and noncompliance.
---------------------------------------------------------------------------

In contrast, a CAFE standard applies to a manufacturer's entire
fleet for a model year. It does not require that a particular
individual vehicle be equipped with any particular equipment or feature
or meet a particular level of fuel economy. It does require that the
manufacturer's fleet, as a whole, comply. Further, although under the
attribute-based approach to setting CAFE standards fuel economy targets
are established for individual vehicles based on their footprints, the
vehicles are not required to comply with those targets. However, as a
practical matter, if a manufacturer chooses to design some vehicles
that fall below their target levels of fuel economy, it will need to
design other vehicles that exceed their targets if the manufacturer's
overall fleet average is to meet the applicable standard.
Thus, under EPCA, there is no such thing as a noncompliant vehicle,
only a noncompliant fleet. No particular vehicle in a noncompliant
fleet is any more, or less, noncompliant than any other vehicle in the
fleet.

C. Development and Feasibility of the Final Standards

1. How was the baseline and reference vehicle fleet developed?
a. Why do the agencies establish a baseline and reference vehicle
fleet?
As also discussed in Section II.B above, in order to determine what
levels of stringency are feasible in future model years, the agencies
must project what vehicles will exist in those model years, and then
evaluate what technologies can feasibly be applied to those vehicles in
order to raise their fuel economy and lower their CO2
emissions. The agencies therefore established a baseline vehicle fleet
representing those vehicles, based on the best available transparent
information. Each agency then developed a separate reference fleet,
accounting (via their respective analytical models) for the effect that
the MY 2011 CAFE standards have on the baseline fleet. This reference
fleet is then used for comparisons of technologies' incremental cost
and effectiveness, as well as for other relevant comparisons in the
rule.
Because NHTSA and EPA have different established practices, the
agencies' rulemaking documents (the Federal Register notice, Joint
Technical Support Document, agency-specific Regulatory Impact Analyses,
and NHTSA Environmental Impact Analysis) have some differences in
terminology. In connection with its first-ever GHG emissions rule under
the CAA, EPA has used the term ``baseline fleet'' to refer to the MY
2008 fleet (i.e., from EPA certification and fuel economy data for MY
2008) prior to adjustment to reflect projected shifts in market
composition. NHTSA, as in recent CAFE rulemakings, refers to the
resultant market forecast, as specified in CAFE model input files (and
corresponding input files for EPA's OMEGA model), as the ``baseline''
fleet. EPA refers to this fleet as the ``reference fleet.'' NHTSA
refers to the ``no action'' standards identified in the EIS (that is,
the MY 2011 standards carried forward through MY 2016) as defining the
``baseline'' scenario, and refers to the fleet to which technologies
have been added in response to these standards as the ``adjusted
baseline'' fleet.\563\ EPA refers to this as the ``final reference
fleet.'' These differences in terminology are summarized in the
following table:
---------------------------------------------------------------------------

\563\ Some manufacturers' baseline fleets (as reflected in the
agencies' market forecast) do not, without applying additional
technology and/or CAFE credits, show compliance with the baseline
standards.

------------------------------------------------------------------------
Fleet description EPA terminology NHTSA terminology
------------------------------------------------------------------------
MY 2008 Fleet with MY 2008 Baseline.......... MY 2008 Fleet
Production Volumes.
MY 2008 Fleet Adjusted to Reference Fleet... Baseline [Market
Reflect Projected Market Shifts. Forecast]
MY 2008 Fleet Adjusted to [Final] Reference Adjusted Baseline
Reflected Projected Market Fleet.
Shifts and Response to MY 2011
CAFE Standards.
------------------------------------------------------------------------

The agencies have retained this mixed terminology in order to
facilitate comparison to past rulemakings. In general, EPA's RIA and
the Joint TSD apply EPA's nomenclature, NHTSA's RIA and EIS apply
NHTSA's nomenclature, and the joint Federal Register notice uses EPA's
nomenclature when focusing on GHG emissions standards, and NHTSA's
nomenclature when focusing on CAFE standards.
b. What data did the agencies use to construct the baseline, and how
did they do so?
As explained in the Technical Support Document (TSD) prepared

[[Page 25559]]

jointly by NHTSA and EPA, both agencies used a baseline vehicle fleet
constructed beginning with EPA fuel economy certification data for the
2008 model year, the most recent model year for which final data is
currently available from manufacturers. These data were used as the
source for MY 2008 production volumes and some vehicle engineering
characteristics, such as fuel economy ratings, engine sizes, numbers of
cylinders, and transmission types.
Some information important for analyzing new CAFE standards is not
contained in the EPA fuel economy certification data. EPA staff
estimated vehicle wheelbase and track widths using data from
Motortrend.com and Edmunds.com. This information is necessary for
estimating vehicle footprint, which is required for the analysis of
footprint-based standards. Considerable additional information
regarding vehicle engineering characteristics is also important for
estimating the potential to add new technologies in response to new
CAFE standards. In general, such information helps to avoid ``adding''
technologies to vehicles that already have the same or a more advanced
technology. Examples include valvetrain configuration (e.g., OHV, SOHC,
DOHC), presence of cylinder deactivation, and fuel delivery (e.g.,
MPFI, SIDI). To the extent that such engineering characteristics were
not available in certification data, EPA staff relied on data published
by Ward's Automotive, supplementing this with information from Internet
sites such as Motortrend.com and Edmunds.com. NHTSA staff also added
some more detailed engineering characteristics (e.g., type of variable
valve timing) using data available from ALLDATA[supreg] Online.
Combined with the certification data, all of this information yielded
the MY 2008 baseline vehicle fleet.
After the baseline was created the next step was to project the
sales volumes for 2011-2016 model years. EPA used projected car and
truck volumes for this period from Energy Information Administration's
(EIA's) 2009 Annual Energy Outlook (AEO).\564\ However, AEO projects
sales only at the car and truck level, not at the manufacturer and
model-specific level, which are needed in order to estimate the effects
new standards will have on individual manufacturers. Therefore, EPA
purchased data from CSM-Worldwide and used their projections of the
number of vehicles of each type predicted to be sold by manufacturers
in 2011-2015.\565\ This provided the year-by-year percentages of cars
and trucks sold by each manufacturer as well as the percentages of each
vehicle segment. The changes between company market share and industry
market segments were most significant from 2011-2014, while for 2014-
2015 the changes were relatively small. Noting this, and lacking a
credible forecast of company and segment shares after 2015, the
agencies assumed 2016 market share and market segments to be the same
as for 2015. Using these percentages normalized to the AEO projected
volumes then provided the manufacturer-specific market share and model-
specific sales for model years 2011-2016.
---------------------------------------------------------------------------

\564\ Available at http://www.eia.doe.gov/oiaf/aeo/index.html
(last accessed March 15, 2010). Specifically, while the total volume
of both cars and trucks was obtained from AEO 2010, the car-truck
split was obtained from AEO 2009. The agencies have also used fuel
price forecasts from AEO 2010. Both agencies regard AEO a credible
source not only of such forecasts, but also of many underlying
forecasts, including forecasts of the size of the future light
vehicle market.
\565\ EPA also considered other sources of similar information,
such as J.D. Powers, and concluded that CSM was more appropriate for
purposes of this rulemaking analysis.
---------------------------------------------------------------------------

The processes for constructing the MY 2008 baseline vehicle fleet
and subsequently adjusting sales volumes to construct the MY 2011-2016
baseline vehicle fleet are presented in detail in Chapter 1 of the
Joint Technical Support Document accompanying today's final rule.
c. How is this different from NHTSA's historical approach and why is
this approach preferable?
As discussed above in Section II.B.4, NHTSA has historically based
its analysis of potential new CAFE standards on detailed product plans
the agency has requested from manufacturers planning to produce light-
duty vehicles for sale in the United States. In contrast, the current
market forecast is based primarily on information sources which are all
either in the public domain or available commercially. There are
advantages to this approach, namely transparency and the potential to
reduce some errors due to manufacturers' misunderstanding of NHTSA's
request for information. There are also disadvantages, namely that the
current market forecast does not represent certain changes likely to
occur in the future vehicle fleet as opposed to the MY 2008 vehicle
fleet, such as vehicles being discontinued and newly introduced. On
balance, however, the agencies have carefully considered these
advantages and disadvantages of using a market forecast derived from
public and commercial sources rather than from manufacturers' product
plans, and conclude that the advantages outweigh the disadvantages.
Although manufacturers did not comment on the agency's proposal to
rely on public and commercial information rather than manufacturers'
confidential product plans when developing a market forecast, those
organizations that did comment on this issue supported this change. The
California Air Resources Board (CARB) and Center for Biological
Diversity (CBD) both commended the resultant increase in transparency.
CARB further indicated that the use of public and commercial
information should produce a better forecast. On the other hand, as
discussed above in Section I, CBD and the Northeast States for
Coordinated Air Use Management (NESCAUM) both raised concerns regarding
the resultant omission of some new vehicle models, and the inclusion of
some vehicles to be discontinued, while CARB suggested that the impact
of these inaccuracies should be minor.
As discussed above in Section II.B.4, while a baseline developed
using publicly and commercially available sources has both advantages
and disadvantages relative to a baseline developed using manufacturers'
product plans, NHTSA has concluded for today's rule that the advantages
outweigh the disadvantages. Today's approach is much more transparent
than the agency's past approach of relying on product plans, and as
discussed in Section II.B.4, any inaccuracies related to new or
discontinued vehicle models should have only a minor impact on the
agency's analysis.
For subsequent rulemakings, NHTSA remains hopeful that
manufacturers will agree to make public their plans for model years
that are very near, so that this information could be incorporated into
analysis available for public review and comment. In any event, because
NHTSA is releasing market inputs used in the agency's analysis of this
final rule, all interested parties can review these inputs fully, as
intended in adopting the transparent approach. More information on the
advantages and disadvantages of the current approach and the agencies'
decision to follow it is available in Section II.B.4.
d. How is this baseline different quantitatively from the baseline that
NHTSA used for the MY 2011 (March 2009) final rule?
As discussed above, the current baseline was developed from
adjusted MY 2008 compliance data and covers MYs 2011-2016, while the
baseline that NHTSA used for the MY 2011 CAFE rule was developed from
confidential

[[Page 25560]]

manufacturer product plans for MY 2011. This section describes, for the
reader's comparison, some of the differences between the current
baseline and the MY 2011 CAFE rule baseline. This comparison provides a
basis for understanding general characteristics and measures of the
difference, in this case, between using publicly (and commercially)
available sources and using manufacturers' confidential product plans.
The current baseline, while developed using the same methods as the
baseline used for MYs 2012-2016 NPRM, reflects updates to the
underlying commercially-available forecast of manufacturer and market
segment shares of the future light vehicle market. These changes are
discussed above in Section II.B.
Estimated vehicle sales:
The sales forecasts, based on the Energy Information
Administration's (EIA's) Annual Energy Outlook 2010 (AEO 2010), used in
the current baseline indicate that the total number of light vehicles
expected to be sold during MYs 2011-2015 is 77 million, or about 15.4
million vehicles annually.\566\ NHTSA's MY 2011 final rule forecast,
based on AEO 2008, of the total number of light vehicles likely to be
sold during MY 2011 through MY 2015 was 83 million, or about 16.6
million vehicles annually. Light trucks are expected to make up 41
percent of the MY 2011 baseline market forecast in the current
baseline, compared to 42 percent of the baseline market forecast in the
MY 2011 final rule. These changes in both the overall size of the light
vehicle market and the relative market shares of passenger cars and
light trucks reflect changes in the economic forecast underlying AEO,
and changes in AEO's forecast of future fuel prices.
---------------------------------------------------------------------------

\566\ Please see Section II.B above and Chapter 1 of the Joint
TSD for more discussion on the agencies' use of AEO 2010 to
determine the sales forecasts for light vehicles during the model
years covered by the rulemaking, as well as the memo available at
Docket No. NHTSA-2009-059-0222.
---------------------------------------------------------------------------

The figures below attempt to demonstrate graphically the difference
between the variation of fuel economy with footprint for passenger cars
under the current baseline and MY 2011 final rule, and for light trucks
under the current baseline and MY 2011 final rule, respectively.
Figures IV.C.1-1 and 1-2 show the variation of fuel economy with
footprint for passenger car models in the current baseline and in the
MY 2011 final rule, while Figures IV.C.1-3 and 1-4 show the variation
of fuel economy with footprint for light truck models in the current
baseline and in the MY 2011 final rule. However, it is difficult to
draw meaningful conclusions by comparing figures from the current
baseline with those of the MY 2011 final rule. In the current baseline
the number of make/models, and their associated fuel economy and
footprint, are fixed and do not vary over time--this is why the number
of data points in the current baseline figures appears smaller as
compared to the number of data points in the MY 2011 final rule
baseline. In contrast, the baseline fleet used in the MY 2011 final
rule varies over time as vehicles (with different fuel economy and
footprint characteristics) are added to and dropped from the product
mix.
BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TR07MY10.023

[[Page 25561]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.024

[[Page 25562]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.025

[[Page 25563]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.026

BILLING CODE 6560-50-C
Estimated manufacturer market shares:
NHTSA's expectations regarding manufacturers' market shares (the
basis for which is discussed below) have also changed since the MY 2011
final rule, given that the agency is relying on different sources of
material for these assumptions as discussed in Section II.B above and
Chapter 1 of the Joint TSD. These changes are reflected below in Table
IV.C.1-1, which shows the agency's sales forecasts for passenger cars
and light trucks under the current baseline and the MY 2011 final
rule.\567\
---------------------------------------------------------------------------

\567\ As explained below, although NHTSA normalized each
manufacturer's overall market share to produce a realistically-sized
fleet, the product mix for each manufacturer that submitted product
plans was preserved. The agency has reviewed manufacturers' product
plans in detail, and understands that manufacturers do not sell the
same mix of vehicles in every model year.

Table IV.C.1-1--Sales Forecasts
[Production for U.S. sale in MY 2011, thousand units]
----------------------------------------------------------------------------------------------------------------
Current baseline MY 2011 Final rule
Manufacturer ---------------------------------------------------------------
Passenger Nonpassenger Passenger Nonpassenger
----------------------------------------------------------------------------------------------------------------
Chrysler........................................ 326 737 707 1,216
Ford............................................ 1,344 792 1,615 1,144
General Motors.................................. 1,249 1,347 1,700 1,844
Honda........................................... 851 585 1,250 470
Hyundai......................................... 382 46 655 221
Kia............................................. 306 88 .............. ..............
Nissan.......................................... 612 331 789 479
Toyota.......................................... 1,356 888 1,405 1,094
Other Asian..................................... 664 246 441 191
European........................................ 833 396 724 190
---------------------------------------------------------------
Total....................................... 7,923 5,458 9,286 6,849
----------------------------------------------------------------------------------------------------------------

[[Page 25564]]

Dual-fueled vehicles:
Manufacturers have also, during and since MY 2008, indicated to the
agency that they intend to sell more dual-fueled or flexible-fuel
vehicles (FFVs) in MY 2011 than indicated in the current baseline of
adjusted MY 2008 compliance data. FFVs create a potential market for
alternatives to petroleum-based gasoline and diesel fuel. For purposes
of determining compliance with CAFE standards, the fuel economy of a
FFV is, subject to limitations, adjusted upward to account for this
potential.\568\ However, NHTSA is precluded from ``taking credit'' for
the compliance flexibility by accounting for manufacturers' ability to
earn and use credits in setting the level of the standards.'' \569\
Some manufacturers plan to produce a considerably greater share of FFVs
than can earn full credit under EPCA. The projected average FFV share
of the market in MY 2011 is 7 percent for the current baseline, versus
17 percent for the MY 2011 final rule. NHTSA notes that in MY 2008 (the
model year providing the vehicle models upon which today's market
forecast is based), the three U.S.-based OEMs produced most of the FFVs
offered for sale in the U.S., yet these OEMs account are projected to
account for a smaller share of the future market in the forecast the
agency has used to develop and analyze today's rule than in the
forecast the agency used to develop and analyze the MY 2011 standards.
---------------------------------------------------------------------------

\568\ See 49 U.S.C. 32905 and 32906.
\569\ 49 U.S.C. 32902(h).
---------------------------------------------------------------------------

Estimated achieved fuel economy levels:
Because manufacturers' product plans also reflect simultaneous
changes in fleet mix and other vehicle characteristics, the
relationship between increased technology utilization and increased
fuel economy cannot be isolated with any certainty. To do so would
require an apples-to-apples ``counterfactual'' fleet of vehicles that
are, except for technology and fuel economy, identical--for example, in
terms of fleet mix and vehicle performance and utility. The current
baseline market forecast shows industry-wide average fuel economy
levels somewhat lower in MY 2011 than shown in the MY 2011 final rule
and the MYs 2012-2016 NPRM. Under the current baseline, average fuel
economy for MY 2011 is 26.4 mpg, versus 26.5 mpg under the baseline in
the MY 2011 final rule, and 26.7 mpg under the baseline in the MYs
2012-2016 NPRM. The 0.3 mpg change relative to the MYs 2012-2016
baseline is the result of changes in manufacturer and market segment
shares of the MY 2011 market.
These differences are shown in greater detail below in Table
IV.C.1-2, which shows manufacturer-specific CAFE levels (not counting
FFV credits that some manufacturers expect to earn) from the current
baseline versus the MY 2011 final rule baseline (from manufacturers'
2008 product plans) for passenger cars and light trucks. Table IV.C.1-3
shows the combined averages of these planned CAFE levels in the
respective baseline fleets. These tables demonstrate that, while the
difference at the industry level is not so large, there are significant
differences in CAFE at the manufacturer level between the current
baseline and the MY 2011 final rule baseline. For example, while
Volkswagen is essentially the same under both, Toyota and Nissan show
increased combined CAFE levels under the current baseline (by 1.9 and
0.7 mpg respectively), while Chrysler, Ford, and GM show decreased
combined CAFE levels under the current baseline (by 1.4, 1.1, and 0.8
mpg, respectively) relative to the MY 2011 final rule baseline.
---------------------------------------------------------------------------

\570\ Again, Kia is not listed in the table for the MY 2011
final rule because it was considered as part of Hyundai for purposes
of that analysis (i.e., Hyundai-Kia).
\571\ Mazda is not listed in the table for the MY 2011 final
rule because it was considered as part of Ford for purposes of that
analysis.
\572\ EPA did not include Ferrari in the current baseline based
on the conclusion that including them would not impact the results,
and therefore Ferrari is not listed in the table for the current
baseline.
\573\ EPA did not include Maserati in the current baseline based
on the conclusion that including them would not impact the results,
and therefore Maserati is not listed in the table for the current
baseline.

Table IV.C.1-2--Current Baseline Planned CAFE Levels in MY 2011 versus MY 2011 Final Rule Planned CAFE Levels
[Passenger and nonpassenger]
----------------------------------------------------------------------------------------------------------------
Current baseline CAFE MY 2011 planned CAFE
levels levels
Manufacturer -----------------------------------------------------
Passenger Nonpassenger Passenger Nonpassenger
----------------------------------------------------------------------------------------------------------------
BMW....................................................... 27.2 23.0 27.0 23.0
Chrysler.................................................. 27.8 21.8 28.2 23.1
Ford...................................................... 28.0 21.0 29.3 22.5
Subaru.................................................... 29.2 26.1 28.6 28.6
General Motors............................................ 28.2 21.2 30.3 21.4
Honda..................................................... 33.5 25.0 32.3 25.2
Hyundai................................................... 32.5 24.3 31.7 26.0
Tata...................................................... 24.6 19.6 24.7 23.9
Kia \570\................................................. 31.7 23.7 ........... ............
Mazda \571\............................................... 30.6 26.0 ........... ............
Daimler................................................... 26.4 21.0 25.2 20.6
Mitsubishi................................................ 29.4 23.6 29.3 26.7
Nissan.................................................... 31.7 21.7 31.3 21.4
Porsche................................................... 26.2 20.0 27.2 20.0
Ferrari \572\............................................. ........... ............ 16.2 ............
Maserati \573\............................................ ........... ............ 18.2 ............
Suzuki.................................................... 30.9 23.3 28.7 24.0
Toyota.................................................... 35.1 23.7 33.2 22.7
Volkswagen................................................ 29.1 20.2 28.5 20.1
-----------------------------------------------------
Total/Average......................................... 30.3 22.2 30.4 22.6
----------------------------------------------------------------------------------------------------------------

[[Page 25565]]

Table IV.C.1-3--Current Baseline Planned CAFE Levels in MY 2011 Versus
MY 2011 Final Rule Planned CAFE Levels (Combined)
------------------------------------------------------------------------
MY 2011
Manufacturer Current Final Rule
baseline baseline
------------------------------------------------------------------------
BMW........................................... 25.0 26.0
Chrysler...................................... 23.3 24.7
Ford.......................................... 24.9 26.0
Subaru........................................ 27.9 28.6
General Motors................................ 24.1 24.9
Honda......................................... 29.5 30.0
Hyundai....................................... 31.3 30.0
Tata.......................................... 21.4 24.4
Kia........................................... 29.5 ...........
Mazda......................................... 29.8 ...........
Daimler....................................... 24.4 23.6
Mitsubishi.................................... 27.4 29.1
Nissan........................................ 27.3 26.6
Porsche....................................... 23.7 22.0
Ferrari....................................... ........... 16.2
Maserati...................................... ........... 18.2
Suzuki........................................ 29.7 27.8
Toyota........................................ 29.5 27.6
Volkswagen.................................... 27.0 27.1
-------------------------
Total/Average............................. 26.4 26.5
------------------------------------------------------------------------

Tables IV.C.1-4 through 1-6 summarize other differences between the
current baseline and manufacturers' product plans submitted to NHTSA in
2008 for the MY 2011 final rule. These tables present average vehicle
footprint, curb weight, and power-to-weight ratios for each
manufacturer represented in the current baseline and of the seven
largest manufacturers represented in the product plan data used in that
rulemaking, and for the overall industry. The tables containing product
plan data do not identify manufacturers by name, and do not present
them in the same sequence.
Tables IV.C.1-4a and 1-4b show that the current baseline reflects a
slight decrease in overall average passenger vehicle size relative to
the manufacturers' plans. This is a reflection of the market segment
shifts underlying the sales forecasts of the current baseline.

Table IV.C.1-4a--Current Baseline Average MY 2011 Vehicle Footprint
[Square feet]
------------------------------------------------------------------------
Manufacturer PC LT Avg.
------------------------------------------------------------------------
BMW.............................. 45.4 49.9 47.5
Chrysler......................... 46.8 52.8 50.9
Daimler.......................... 47.1 53.3 49.0
Ford............................. 46.3 56.1 49.9
General Motors................... 46.4 58.2 52.5
Honda............................ 44.3 49.1 46.3
Hyundai.......................... 44.4 48.7 44.8
Kia.............................. 45.2 51.0 46.5
Mazda............................ 44.4 47.3 44.9
Mitsubishi....................... 43.8 46.5 44.6
Nissan........................... 45.3 53.9 48.3
Porsche.......................... 38.6 51.0 42.8
Subaru........................... 43.1 46.2 44.3
Suzuki........................... 40.8 47.2 41.6
Tata............................. 50.3 47.8 48.8
Toyota........................... 44.0 53.0 47.6
Volkswagen....................... 43.5 52.6 45.1
--------------------------------------
Industry Average............. 45.2 53.5 48.6
------------------------------------------------------------------------

Table IV.C.1-4b--MY 2011 Final Rule Average Planned MY 2011 Vehicle
Footprint
[Square feet]
------------------------------------------------------------------------
PC LT Avg.
------------------------------------------------------------------------
Manufacturer 1................... 46.7 58.5 52.8
Manufacturer 2................... 46.0 50.4 47.1
Manufacturer 3................... 44.9 52.8 48.4
Manufacturer 4................... 45.4 55.8 49.3
Manufacturer 5................... 45.2 57.5 50.3
Manufacturer 6................... 48.5 54.7 52.4
Manufacturer 7................... 45.1 49.9 46.4
--------------------------------------
Industry Average............. 45.6 55.1 49.7
------------------------------------------------------------------------

Tables IV.C.1-5a and 1-5b show that the current baseline reflects a
decrease in overall average vehicle weight relative to the
manufacturers' plans. As above, this is most likely a reflection of the
market segment shifts underlying the sales forecasts of the current
baseline.

[[Page 25566]]

Table IV.C.1-5a--Current Baseline Average MY 2011 Vehicle Curb Weight
[Pounds]
------------------------------------------------------------------------
Manufacturer PC LT Avg.
------------------------------------------------------------------------
BMW.............................. 3,535 4,648 4,055
Chrysler......................... 3,572 4,469 4,194
Daimler.......................... 3,583 5,127 4,063
Ford............................. 3,526 4,472 3,877
General Motors................... 3,528 4,978 4,281
Honda............................ 3,040 4,054 3,453
Hyundai.......................... 3,014 4,078 3,129
Kia.............................. 3,035 4,007 3,252
Mazda............................ 3,258 3,803 3,348
Mitsubishi....................... 3,298 3,860 3,468
Nissan........................... 3,251 4,499 3,689
Porsche.......................... 3,159 4,906 3,760
Subaru........................... 3,176 3,470 3,391
Suzuki........................... 2,842 3,843 2,965
Tata............................. 3,906 5,171 4,627
Toyota........................... 3,109 4,321 3,589
Volkswagen....................... 3,445 5,672 3,839
--------------------------------------
Industry Average............. 3,313 4,499 3,797
------------------------------------------------------------------------

Table IV.C.1-5b--MY 2011 Final Rule Average Planned MY 2011 Vehicle Curb
Weight
[Pounds]
------------------------------------------------------------------------
PC LT Avg.
------------------------------------------------------------------------
Manufacturer 1................... 3,197 4,329 3,692
Manufacturer 2................... 3,691 4,754 4,363
Manufacturer 3................... 3,293 4,038 3,481
Manufacturer 4................... 3,254 4,191 3,510
Manufacturer 5................... 3,547 5,188 4,401
Manufacturer 6................... 3,314 4,641 3,815
Manufacturer 7................... 3,345 4,599 3,865
--------------------------------------
Industry Average............. 3,380 4,687 3,935
------------------------------------------------------------------------

Tables IV.C.1-6a and IV.C.1-6b show that the current baseline
reflects a decrease in average performance relative to that of the
manufacturers' product plans. This decreased performance is most likely
a reflection of the market segment shifts underlying the sales
forecasts of the current baseline, that is, an assumed shift away from
higher performance vehicles.

Table IV.C.1-6a--Current Baseline Average MY 2011 Vehicle Power-to-
Weight Ratio
[hp/lb]
------------------------------------------------------------------------
Manufacturer PC LT Avg.
------------------------------------------------------------------------
BMW.............................. 0.072 0.061 0.067
Chrysler......................... 0.055 0.052 0.053
Daimler.......................... 0.068 0.056 0.064
Ford............................. 0.058 0.054 0.056
General Motors................... 0.057 0.056 0.056
Honda............................ 0.056 0.054 0.056
Hyundai.......................... 0.052 0.055 0.052
Kia.............................. 0.050 0.056 0.051
Mazda............................ 0.052 0.055 0.052
Mitsubishi....................... 0.053 0.056 0.054
Nissan........................... 0.059 0.057 0.058
Porsche.......................... 0.105 0.073 0.094
Subaru........................... 0.060 0.056 0.058
Suzuki........................... 0.049 0.062 0.051
Tata............................. 0.077 0.057 0.065
Toyota........................... 0.053 0.062 0.056
Volkswagen....................... 0.057 0.052 0.056
--------------------------------------
Industry Average............. 0.057 0.056 0.056
------------------------------------------------------------------------

[[Page 25567]]

Table IV.C.1-6b--MY 2011 Final Rule Average Planned MY 2011 Vehicle
Power-to-Weight Ratio
[hp/lb]
------------------------------------------------------------------------
PC LT Avg.
------------------------------------------------------------------------
Manufacturer 1................... 0.065 0.058 0.060
Manufacturer 2................... 0.061 0.065 0.062
Manufacturer 3................... 0.053 0.059 0.056
Manufacturer 4................... 0.060 0.058 0.059
Manufacturer 5................... 0.060 0.057 0.059
Manufacturer 6................... 0.063 0.065 0.065
Manufacturer 7................... 0.053 0.055 0.053
--------------------------------------
Industry Average............. 0.060 0.059 0.060
------------------------------------------------------------------------

As discussed above, the agencies' market forecast for MY 2012-2016
holds the performance and other characteristics of individual vehicle
models constant, adjusting the size and composition of the fleet from
one model year to the next.
Refresh and redesign schedules (for application in NHTSA's
modeling):
Expected model years in which each vehicle model will be redesigned
or freshened constitute another important aspect of NHTSA's market
forecast. As discussed in Section IV.C.2.c below, NHTSA's analysis
supporting the current rulemaking times the addition of nearly all
technologies to coincide with either a vehicle redesign or a vehicle
freshening. Product plans submitted to NHTSA preceding the MY 2011
final rule contained manufacturers' estimates of vehicle redesign and
freshening schedules and NHTSA's estimates of the timing of the five-
year redesign cycle and the two- to three-year refresh cycle were made
with reference to those plans. In the current baseline, in contrast,
estimates of the timing of the refresh and redesign cycles were based
on historical dates--i.e., counting forward from known redesigns
occurring in or prior to MY 2008 for each vehicle in the fleet and
assigning refresh and redesign years accordingly. After applying these
estimates, the shares of manufacturers' passenger car and light truck
estimated to be redesigned in MY 2011 were as summarized below for the
current baseline and the MY 2011 final rule. Table IV.C.1-7 below shows
the percentages of each manufacturer's fleets expected to be redesigned
in MY 2011 for the current baseline. Table IV.C.1-8 presents
corresponding estimates from the market forecast used by NHTSA in the
analysis supporting the MY 2011 final rule (again, to protect
confidential information, manufacturers are not identified by name).

Table IV.C.1-7--Current Baseline, Share of Fleet Redesigned in MY 2011
------------------------------------------------------------------------
PC LT Avg.
Manufacturer (percent) (percent) (percent)
------------------------------------------------------------------------
BMW.............................. 32 37 34
Chrysler......................... 0 13 9
Daimler.......................... 0 0 0
Ford............................. 12 8 11
General Motors................... 17 3 9
Honda............................ 29 26 28
Hyundai.......................... 26 0 23
Kia.............................. 38 83 48
Mazda............................ 0 0 0
Mitsubishi....................... 0 59 18
Nissan........................... 5 25 12
Porsche.......................... 0 100 34
Subaru........................... 0 42 16
Suzuki........................... 4 21 6
Tata............................. 28 100 69
Toyota........................... 5 15 9
Volkswagen....................... 16 0 13
--------------------------------------
Industry Average............. 13 15 14
------------------------------------------------------------------------

[[Page 25568]]

Table IV.C.1-8--MY 2011 Final Rule, Share of Fleet Redesigned in MY 2011
------------------------------------------------------------------------
PC LT Avg.
(percent) (percent) (percent)
------------------------------------------------------------------------
Manufacturer 1................... 19 0 11
Manufacturer 2................... 34 27 29
Manufacturer 3................... 5 0 3
Manufacturer 4................... 7 0 5
Manufacturer 5................... 19 0 11
Manufacturer 6................... 34 28 33
Manufacturer 7................... 27 28 28
--------------------------------------
Overall...................... 20 9 15
------------------------------------------------------------------------

We continue, therefore, to estimate that manufacturers' redesigns
will not be uniformly distributed across model years. This is in
keeping with standard industry practices, and reflects what
manufacturers actually do--NHTSA has observed that manufacturers in
fact do redesign more vehicles in some years than in others. NHTSA
staff have closely examined manufacturers' planned redesign schedules,
contacting some manufacturers for clarification of some plans, and
confirmed that these plans remain unevenly distributed over time. For
example, although Table IV.C.1-8 shows that NHTSA expects Company 2 to
redesign 34 percent of its passenger car models in MY 2011, current
information indicates that this company will then redesign only (a
different) 10 percent of its passenger cars in MY 2012. Similarly,
although Table IV.C.1-8 shows that NHTSA expects four of the largest
seven light truck manufacturers to redesign virtually no light truck
models in MY 2011, current information also indicates that these four
manufacturers will redesign 21-49 percent of their light trucks in MY
2012.
e. How does manufacturer product plan data factor into the baseline
used in this rule?
As discussed in Section II.B.5 above, while the agencies received
updated product plans in Spring and Fall 2009 in response to NHTSA's
requests, the baseline data used in this final rule is not informed by
these product plans, except with respect to specific engineering
characteristics (e.g., GVWR) of some MY 2008 vehicle models, because
these product plans contain confidential business information that the
agencies are legally required to protect from disclosure, and because
the agencies have concluded that, for purposes of this final rule, a
transparent baseline is preferable.
For the NPRM, NHTSA conducted a separate analysis that did make use
of these product plans. NHTSA performed this separate analysis for
purposes of comparison only. For today's final rule NHTSA used the
publicly available baseline for all analysis related to the development
and evaluation of the new CAFE standards. As discussed above in Section
II.B.4, while a baseline developed using publicly and commercially
available sources has both advantages and disadvantages relative to a
baseline developed using manufacturers' product plans, NHTSA has
concluded for today's rule that the advantages outweigh the
disadvantages. NHTSA plans to consider these advantages and
disadvantages further in connection with future rulemakings, taking
into account changes in the market, changes in the scope and quality of
publicly and commercially available data, and any changes in
manufacturers' willingness to make some product planning information
publicly available.
2. How were the technology inputs developed?
As discussed above in Section II.E, for developing the technology
inputs for the MY 2012-2016 CAFE and GHG standards, the agencies
primarily began with the technology inputs used in the MY 2011 CAFE
final rule and in the July 2008 EPA ANPRM, and then reviewed, as
requested by President Obama in his January 26 memorandum, the
technology assumptions that NHTSA used in setting the MY 2011 standards
and the comments that NHTSA received in response to its May 2008 Notice
of Proposed Rulemaking, as well as the comments received to the NPRM
for this rule. In addition, the agencies supplemented their review with
updated information from the FEV tear-down studies contracted by EPA,
more current literature, new product plans and from EPA certification
testing. More detail is available regarding how the agencies developed
the technology inputs for this final rule above in Section II.E, in
Chapter 3 of the Joint TSD, and in Section V of NHTSA's FRIA.
a. What technologies does NHTSA consider?
Section II.E.1 above describes the fuel-saving technologies
considered by the agencies that manufacturers could use to improve the
fuel economy of their vehicles during MYs 2012-2016. The majority of
the technologies described in this section are readily available, well
known, and could be incorporated into vehicles once production
decisions are made. As discussed, the technologies considered fall into
five broad categories: engine technologies, transmission technologies,
vehicle technologies, electrification/accessory technologies, and
hybrid technologies. Table IV.C.2-1 below lists all the technologies
considered and provides the abbreviations used for them in the Volpe
model,\574\ as well as their year of availability, which for purposes
of NHTSA's analysis means the first model year in the rulemaking period
that the Volpe model is allowed to apply a technology to a
manufacturer's fleet.\575\ Year of availability recognizes that
technologies must achieve a level of technical viability before they
can be implemented in the Volpe model, and are thus a means of
constraining technology use until such time as it is considered to be
technologically feasible. For a more detailed description of each
technology and their costs and effectiveness, we refer the reader to
Chapter 3 of the Joint TSD and Section V of NHTSA's FRIA.
---------------------------------------------------------------------------

\574\ The abbreviations are used in this section both for
brevity and for the reader's reference if they wish to refer to the
expanded decision trees and the model input and output sheets, which
are available in Docket No. NHTSA-2009-0059-0156 and on NHTSA's Web
site.
\575\ A date of 2011 means the technology can be applied in all
model years, while a date of 2014 means the technology can only be
applied in model years 2014 through 2016.

[[Page 25569]]

Table IV.C.2-1--List of Technologies in NHTSA's Analysis
------------------------------------------------------------------------
Technology Model abbreviation Year available
------------------------------------------------------------------------
Low Friction Lubricants........... LUB................. 2011
Engine Friction Reduction......... EFR................. 2011
VVT--Coupled Cam Phasing (CCP) on CCPS................ 2011
SOHC.
Discrete Variable Valve Lift DVVLS............... 2011
(DVVL) on SOHC.
Cylinder Deactivation on SOHC..... DEACS............... 2011
VVT--Intake Cam Phasing (ICP)..... ICP................. 2011
VVT--Dual Cam Phasing (DCP)....... DCP................. 2011
Discrete Variable Valve Lift DVVLD............... 2011
(DVVL) on DOHC.
Continuously Variable Valve Lift CVVL................ 2011
(CVVL).
Cylinder Deactivation on DOHC..... DEACD............... 2011
Cylinder Deactivation on OHV...... DEACO............... 2011
VVT--Coupled Cam Phasing (CCP) on CCPO................ 2011
OHV.
Discrete Variable Valve Lift DVVLO............... 2011
(DVVL) on OHV.
Conversion to DOHC with DCP....... CDOHC............... 2011
Stoichiometric Gasoline Direct SGDI................ 2011
Injection (GDI).
Combustion Restart................ CBRST............... 2014
Turbocharging and Downsizing...... TRBDS............... 2011
Exhaust Gas Recirculation (EGR) EGRB................ 2013
Boost.
Conversion to Diesel following DSLC................ 2011
CBRST.
Conversion to Diesel following DSLT................ 2011
TRBDS.
6-Speed Manual/Improved Internals. 6MAN................ 2011
Improved Auto. Trans. Controls/ IATC................ 2011
Externals.
Continuously Variable Transmission CVT................. 2011
6/7/8-Speed Auto. Trans with NAUTO............... 2011
Improved Internals.
Dual Clutch or Automated Manual DCTAM............... 2011
Transmission.
Electric Power Steering........... EPS................. 2011
Improved Accessories.............. IACC................ 2011
12V Micro-Hybrid.................. MHEV................ 2011
Belt Integrated Starter Generator. BISG................ 2011
Crank Integrated Starter Generator CISG................ 2011
Power Split Hybrid................ PSHEV............... 2011
2-Mode Hybrid..................... 2MHEV............... 2011
Plug-in Hybrid.................... PHEV................ 2011
Mass Reduction 1 (1.5%)........... MS1................. 2011
Mass Reduction 2 (3.5%-8.5%)...... MS2................. 2014
Low Rolling Resistance Tires...... ROLL................ 2011
Low Drag Brakes................... LDB................. 2011
Secondary Axle Disconnect 4WD..... SAX................. 2011
Aero Drag Reduction............... AERO................ 2011
------------------------------------------------------------------------

For purposes of this final rule and as discussed in greater detail
in the Joint TSD, NHTSA and EPA carefully reviewed the list of
technologies used in the agency's analysis for the MY 2011 final rule.
NHTSA and EPA concluded that the considerable majority of technologies
were correctly defined and continued to be appropriate for use in the
analysis supporting the final standards. However, some refinements were
made as discussed in the NPRM.\576\ Additionally, the following
refinements were made for purposes of the final rule.
---------------------------------------------------------------------------

\576\ 74 FR at 49655-56 (Sept. 28, 2009).
---------------------------------------------------------------------------

Specific to its modeling, NHTSA has revised two technologies used
in the final rule analysis from those considered in the NPRM. These
revisions were based on comments received in response to the NPRM and
the identification of area to improve accuracy. In the NPRM, a diesel
engine option (DSLT or DSLC) was not available for small vehicles
because it did not appear to be a cost-effective option. However, based
on comments received in response to the NPRM, the agency added a diesel
engine option for small vehicles. Additionally, in the NPRM, the mass
reduction/material substitution technology, MS1, assumed engine
downsizing. However, for purposes of the final rule, engine downsizing
is no longer assumed for MS1, thus slightly lowering the effectiveness
estimate to better reflect how manufacturers might implement small
amounts of mass reduction/material substitution. Chapter 3 of the Joint
TSD and Section V of NHTSA's FRIA provide a more detailed explanation
of these revisions.
b. How did NHTSA determine the costs and effectiveness of each of these
technologies for use in its modeling analysis?
Building on NHTSA's estimates developed for the MY 2011 CAFE final
rule and EPA's Advanced Notice of Proposed Rulemaking, which relied on
EPA's 2008 Staff Technical Report,\577\ the agencies took a fresh look
at technology cost and effectiveness values and incorporated additional
FEV tear-down study results for purposes of this final rule. This joint
work is reflected in Chapter 3 of the Joint TSD and in Section II of
this preamble, as summarized below. For more detailed information on
the effectiveness and cost of fuel-saving technologies, please refer to
Chapter 3 of the Joint TSD and Section V of NHTSA's FRIA. NHTSA and EPA
are confident that the thorough review conducted for purposes of this
final rule led to the best available conclusions regarding technology
costs and effectiveness estimates for the current rulemaking and
resulted in excellent consistency between the agencies' respective
analyses for

[[Page 25570]]

developing the CAFE and CO2 standards.
---------------------------------------------------------------------------

\577\ EPA Staff Technical Report: Cost and Effectiveness
Estimates of Technologies Used to Reduce Light-Duty Vehicle Carbon
Dioxide Emissions. EPA420-R-08-008, March 2008. Available at Docket
No. NHTSA-2009-0059-0027.
---------------------------------------------------------------------------

Generally speaking, while NHTSA and EPA found that much of the cost
information used in NHTSA's MY 2011 final rule and EPA's 2008 Staff
Report was consistent to a great extent, the agencies, in reconsidering
information from many sources revised several component costs of
several major technologies for purposes of the NRPM: mild and strong
hybrids, diesels, SGDI, and Valve Train Lift Technologies. In addition,
based on FEV tear-down studies, the costs for turbocharging/downsizing,
6-, 7-, 8-speed automatic transmissions, and dual clutch transmissions
were revised for this final rule. These revisions are discussed at
length in the Joint TSD and in NHTSA's FRIA.
Most effectiveness estimates used in both the MY 2011 final rule
and the 2008 EPA Staff Report were determined to be accurate and were
carried forward without significant change into this rulemaking. When
NHTSA and EPA's estimates for effectiveness diverged slightly due to
differences in how the agencies apply technologies to vehicles in their
respective models, we report the ranges for the effectiveness values
used in each model. For purposes of the final rule analysis, NHTSA made
only a couple of changes to the effectiveness estimates. Specifically,
in reviewing the NPRM effectiveness estimates for this final rule NHTSA
discovered that the DCTAM effectiveness value for Subcompact and
Compact subclasses was incorrect; the (lower) wet clutch effectiveness
estimate had been used instead of the intended (higher) dry clutch
estimate for these vehicle classes.\578\ Thus, NHTSA corrected these
effectiveness estimates. Additionally, as discussed above, the
effectiveness estimate for MS1 was revised (lowered) to better
represent the impact of reducing mass at a refresh. For much more
information on the costs and effectiveness of individual technologies,
we refer the reader to Chapter 3 of the Joint TSD and Section V of
NHTSA's FRIA.
---------------------------------------------------------------------------

\578\ ``Dry clutch'' DCTAMs and ``wet clutch'' DCTAMs have
different characteristics and different uses. A dry clutch DCTAM is
more efficient and less expensive than a wet clutch DCTAM, which
requires a wet-clutch-type hydraulic system to cool the clutches.
However, without a cooling system, a dry clutch DCTAM has a lower
torque capacity. Dry clutch DCTAMs are thus ideal for smaller
vehicles with lower torque ratings, like those in the Subcompact and
Compact classes, while wet clutch DCTAMs would be more appropriate
for, e.g., larger trucks. Thus, it is appropriate to distinguish
accordingly in DCTAM effectiveness between subclasses.
---------------------------------------------------------------------------

As a general matter, NHTSA received relatively few comments related
to technology cost and effectiveness estimates as compared to the
number received on these issues in previous CAFE rulemakings. The
California Air Resources Board (CARB) generally agreed with cost
estimates used in the NPRM analysis. NHTSA also received comments from
the Aluminum Association, General Motors, Honeywell, International
Council on Clean Transportation (ICCT), Manufacturers of Emission
Controls Association (MECA), Motor and Equipment Manufacturers
Association (MEMA) and the New Jersey Department of Environmental
Protection related to cost and effectiveness estimates for specific
technologies, including but not limited to hybrids, diesels,
turbocharging and downsizing, and mass reduction/material substitution.
A detailed description of these comments and NHTSA's responses can be
found in Section V of NHTSA's FRIA.
NHTSA notes that, in developing technology cost and effectiveness
estimates, the agencies have made every effort to hold constant aspects
of vehicle performance and utility typically valued by consumers, such
as horsepower, carrying capacity, and towing and hauling capacity. For
example, NHTSA includes in its analysis technology cost and
effectiveness estimates that are specific to performance passenger cars
(i.e., sports cars), as compared to non-performance passenger cars.
NHTSA sought comment on the extent to which commenters believed that
the agencies have been successful in holding constant these elements of
vehicle performance and utility in developing the technology cost and
effectiveness estimates, but received relatively little in response.
NHTSA thus concludes that commenters had no significant issues with its
approach for purposes of this rulemaking, but the agency will continue
to analyze this issue going forward.
Additionally, NHTSA notes that the technology costs included in
this final rule take into account only those associated with the
initial build of the vehicle. The agencies sought comment on the
additional lifetime costs, if any, associated with the implementation
of advanced technologies, including warranty, maintenance and
replacement costs, such as the replacement costs for low rolling
resistance tires, low friction lubricants, and hybrid batteries, and
maintenance costs for diesel aftertreatment components, but received no
responses. The agency will continue to examine this issue closely for
subsequent rulemakings, particularly as manufacturers turn increasingly
to even more advanced technologies in the future that may have more
significant lifetime costs.
The tables below provide examples of the incremental cost and
effectiveness estimates employed by the agency in developing this final
rule, according to the decision trees used in the Volpe modeling
analysis. Thus, the effectiveness and cost estimates are not absolute
to a single reference vehicle, but are incremental to the technology or
technologies that precede it.

Table IV.C.2-2--Technology Effectiveness Estimates Employed in the Volpe Model for Certain Technologies
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Perform. Perform. Perform.
Subcomp. Compact Midsize Large car subcomp. compact midsize Perform. Minivan Small LT Midsize Large LT
car car car car car car large car LT LT
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
VEHICLE TECHNOLOGY INCREMENTAL FUEL CONSUMPTION REDUCTION (-%)
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Low Friction Lubricants..................................... 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
VVT--Dual Cam Phasing (DCP)................................. 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0
Discrete Variable Valve Lift (DVVL) on DOHC................. 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0 1.0-3.0
Cylinder Deactivation on OHV................................ n.a. n.a. n.a. 3.9-5.5 n.a. 3.9-5.5 3.9-5.5 3.9-5.5 3.9-5.5 n.a. 3.9-5.5 3.9-5.5

[[Page 25571]]

Stoichiometric Gasoline Direct Injection (GDI).............. 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0
Turbocharging and Downsizing................................ 4.2-4.8 4.2-4.8 4.2-4.8 1.8-1.9 4.2-4.8 1.8-1.9 1.8-1.9 1.8-1.9 1.8-1.9 4.2-4.8 1.8-1.9 1.8-1.9
6/7/8-Speed Auto. Trans with Improved Internals............. 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4 1.4-3.4
Electric Power Steering..................................... 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0 1.0-2.0
12V Micro-Hybrid............................................ 2.0-3.0 2.0-3.0 2.0-3.0 2.5-3.5 2.0-3.0 2.5-3.5 2.5-3.5 3.0-4.0 2.5-3.5 2.0-3.0 2.5-3.5 n.a.
Crank mounted Integrated Starter Generator.................. 8.6-8.9 8.6-8.9 8.6-8.9 8.7-8.9 8.6-8.9 8.7-8.9 8.7-8.9 8.7-8.9 8.7-8.9 8.6-8.9 8.7-8.9 14.1-16.3
Power Split Hybrid.......................................... 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 6.3-12.4 n.a.
Aero Drag Reduction......................................... 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Table IV.C.2-3--Technology Cost Estimates Employed in the Volpe Model for Certain Technologies
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Perform. Perform. Perform.
Subcomp. Compact Midsize Large car subcomp. compact midsize Perform. Minivan Small LT Midsize Large LT
car car car car car car large car LT LT
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
VEHICLE TECHNOLOGY ICM COSTS PER VEHICLE ($)
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Nominal baseline engine (for cost purpose).................. (*) (*) (*) V6 (*) V6 V6 V8 V6 (*) V6 V8
Low Friction Lubricants..................................... 3 3 3 3 3 3 3 3 3 3 3 3
VVT--Dual Cam Phasing (DCP)................................. 38 38 38 82 38 82 82 82 82 38 82 82
Discrete Variable Valve Lift (DVVL) on DOHC................. 142 142 142 206 142 206 206 294 206 142 206 294
Cylinder Deactivation on OHV................................ n.a. n.a. n.a. 168 n.a. 168 168 192 168 n.a. 168 192
Stoichiometric Gasoline Direct Injection (GDI).............. 236 236 236 342 236 342 342 392 342 236 342 392
Turbocharging and Downsizing................................ 445 445 445 325 445 325 325 919 325 445 325 919
6/7/8-Speed Auto. Trans with Improved Internals............. 112 112 112 112 112-214 112-214 112-214 112-214 112-214 112 112-214 112-214
Electric Power Steering..................................... 106 106 106 106 106 106 106 106 106 106 106 106
12V Micro-Hybrid............................................ 288 311 342 367 314 337 372 410 337 325 376 n.a.
Crank mounted Integrated Starter Generator.................. 2,791 3,107 3,319 3,547 2,839 3,149 3,335 3,571 3,149 3,141 3,611 5,124
Power Split Hybrid.......................................... 1,600 2,133 2,742 3,261 3,661 4,018 5,287 6,723 4,018 2,337 3,462 n.a.
Aero Drag Reduction......................................... 48 48 48 48 48 48 48 48 48 48 48 48
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
* Inline 4.

[[Page 25572]]

c. How does NHTSA use these assumptions in its modeling analysis?
NHTSA relies on several inputs and data files to conduct the
compliance analysis using the Volpe model, as discussed further below
and in Section V of the FRIA. For the purposes of applying
technologies, the Volpe model primarily uses two data files, one that
contains data on the vehicles expected to be manufactured in the model
years covered by the rulemaking and identifies the appropriate stage
within the vehicle's life-cycle for the technology to be applied, and
one that contains data/parameters regarding the available technologies
the model can apply. These inputs are discussed below.
As discussed above, the Volpe model begins with an initial state of
the domestic vehicle market, which in this case is the market for
passenger cars and light trucks to be sold during the period covered by
the final standards. The vehicle market is defined on a model-by-model,
engine-by-engine, and transmission-by-transmission basis, such that
each defined vehicle model refers to a separately defined engine and a
separately defined transmission.
For the current standards, which cover MYs 2012-2016, the light-
duty vehicle (passenger car and light truck) market forecast was
developed jointly by NHTSA and EPA staff using MY 2008 CAFE compliance
data. The MY 2008 compliance data includes about 1,100 vehicle models,
about 400 specific engines, and about 200 specific transmissions, which
is a somewhat lower level of detail in the representation of the
vehicle market than that used by NHTSA in recent CAFE analyses--
previous analyses would count a vehicle as ``new'' in any year when
significant technology differences are made, such as at a
redesign.\579\ However, within the limitations of information that can
be made available to the public, it provides the foundation for a
realistic analysis of manufacturer-specific costs and the analysis of
attribute-based CAFE standards, and is much greater than the level of
detail used by many other models and analyses relevant to light-duty
vehicle fuel economy.\580\
---------------------------------------------------------------------------

\579\ The market file for the MY 2011 final rule, which included
data for MYs 2011-2015, had 5500 vehicles, about 5 times what we are
using in this analysis of the MY 2008 certification data.
\580\ Because CAFE standards apply to the average performance of
each manufacturer's fleet of cars and light trucks, the impact of
potential standards on individual manufacturers cannot be credibly
estimated without analysis of the fleets that manufacturers can be
expected to produce in the future. Furthermore, because required
CAFE levels under an attribute-based CAFE standard depend on
manufacturers' fleet composition, the stringency of an attribute-
based standard cannot be predicted without performing analysis at
this level of detail.
---------------------------------------------------------------------------

In addition to containing data about each vehicle, engine, and
transmission, this file contains information for each technology under
consideration as it pertains to the specific vehicle (whether the
vehicle is equipped with it or not), the estimated model year the
vehicle is undergoing redesign, and information about the vehicle's
subclass for purposes of technology application. In essence, the model
considers whether it is appropriate to apply a technology to a vehicle.

Is a vehicle already equipped, or can it not be equipped, with a
particular technology?

The market forecast file provides NHTSA the ability to identify, on
a technology by technology basis, which technologies may already be
present (manufactured) on a particular vehicle, engine, or
transmission, or which technologies are not applicable (due to
technical considerations) to a particular vehicle, engine, or
transmission. These identifications are made on a model-by-model,
engine-by-engine, and transmission-by-transmission basis. For example,
if the market forecast file indicates that Manufacturer X's Vehicle Y
is manufactured with Technology Z, then for this vehicle Technology Z
will be shown as used. Additionally, NHTSA has determined that some
technologies are only suitable or unsuitable when certain vehicle,
engine, or transmission conditions exist. For example, secondary axle
disconnect is only suitable for 4WD vehicles, and cylinder deactivation
is unsuitable for any engine with fewer than 6 cylinders, while CVTs
can only be applied to unibody vehicles. Similarly, comments received
to the 2008 NPRM indicated that cylinder deactivation could not likely
be applied to vehicles equipped with manual transmissions during the
rulemaking timeframe, due primarily to the cylinder deactivation system
not being able to anticipate gear shifts. The Volpe model employs
``engineering constraints'' to address issues like these, which are a
programmatic method of controlling technology application that is
independent of other constraints. Thus, the market forecast file would
indicate that the technology in question should not be applied to the
particular vehicle/engine/transmission (i.e., is unavailable). Since
multiple vehicle models may be equipped with an engine or transmission,
this may affect multiple models. In using this aspect of the market
forecast file, NHTSA ensures the Volpe model only applies technologies
in an appropriate manner, since before any application of a technology
can occur, the model checks the market forecast to see if it is either
already present or unavailable.
In response to the NPRM, NHTSA received comments from GM that
included a description of technical considerations, concerns,
limitations and risks that need to be considered when implementing
turbocharging and downsizing technologies on full size trucks. These
include concerns related to engine knock, drivability, control of boost
pressure, packaging complexity, enhanced cooling for vehicles that are
designed for towing or hauling, and noise, vibration and harshness.
NHTSA judges that the expressed technical considerations, concerns,
limitations and risks are well recognized within the industry and it is
standard industry practice to address each during the design and
development phases of applying turbocharging and downsizing
technologies. Cost and effectiveness estimates used in the final rule
are based on analysis that assumes each of these factors is addressed
prior to production implementation of the technologies. In comments
related to full size trucks, GM commented that potential to address
knock limit concerns through various alternatives, which include use of
higher octane premium fuel and/or the addition of a supplemental
ethanol injection system. For this rulemaking, NHTSA has not assumed
that either of these approaches is implemented to address knock limit
concerns, and these technologies are not included in assessment of
turbocharging and downsizing feasibility, cost or effectiveness.\581\
In addition, NHTSA has received confidential business information from
a manufacturer that supports that turbocharging and downsizing is
feasible on a full size truck product during the rulemaking period.

\581\ Note that for one of the teardown analysis cost studies of
turbocharging and downsizing conducted by FEV, in which a 2.4L I4
DOHC naturally aspirated engine was replaced by a 1.6L I4 DOHC SGDI
turbocharged engine, the particular 1.6L turbocharged engine chosen
for the study was a premium octane fuel engine. For this rulemaking,
NHTSA intends that a turbocharged and downsized engine achieve
comparable performance to a baseline engine without requiring
premium octane fuel. For the FEV study of the 1.6L turbocharged
engine, this could be achieved through the specification of an
engine with a displacement of slightly greater than 1.6L. NHTSA
judges that a slightly larger engine would have small effect on the
overall cost analysis used in this rulemaking. For all other
teardown studies conducted by FEV, both the naturally aspirated
engine and the replacement turbocharged and downsized engine were
specified to use regular octane fuel.

---------------------------------------------------------------------------

[[Page 25573]]

---------------------------------------------------------------------------
Is a vehicle being redesigned or refreshed?

Manufacturers typically plan vehicle changes to coincide with
certain stages of a vehicle's life cycle that are appropriate for the
change, or in this case the technology being applied. In the automobile
industry there are two terms that describe when technology changes to
vehicles occur: Redesign and refresh (i.e., freshening). Vehicle
redesign usually refers to significant changes to a vehicle's
appearance, shape, dimensions, and powertrain. Redesign is
traditionally associated with the introduction of ``new'' vehicles into
the market, often characterized as the ``next generation'' of a
vehicle, or a new platform. Vehicle refresh usually refers to less
extensive vehicle modifications, such as minor changes to a vehicle's
appearance, a moderate upgrade to a powertrain system, or small changes
to the vehicle's feature or safety equipment content. Refresh is
traditionally associated with mid-cycle cosmetic changes to a vehicle,
within its current generation, to make it appear ``fresh.'' Vehicle
refresh generally occurs no earlier than two years after a vehicle
redesign, or at least two years before a scheduled redesign. For the
majority of technologies discussed today, manufacturers will only be
able to apply them at a refresh or redesign, because their application
would be significant enough to involve some level of engineering,
testing, and calibration work.\582\
---------------------------------------------------------------------------

\582\ For example, applying material substitution through weight
reduction, or even something as simple as low rolling-resistance
tires, to a vehicle will likely require some level of validation and
testing to ensure that the vehicle may continue to be certified as
compliant with NHTSA's Federal Motor Vehicle Safety Standards
(FMVSS). Weight reduction might affect a vehicle's crashworthiness;
low rolling-resistance tires might change a vehicle's braking
characteristics or how it performs in crash avoidance tests.
---------------------------------------------------------------------------

Some technologies (e.g., those that require significant revision)
are nearly always applied only when the vehicle is expected to be
redesigned, like turbocharging and engine downsizing, or conversion to
diesel or hybridization. Other technologies, like cylinder
deactivation, electric power steering, and aerodynamic drag reduction
can be applied either when the vehicle is expected to be refreshed or
when it is expected to be redesigned, while a few others, like low
friction lubricants, can be applied at any time, regardless of whether
a refresh or redesign event is conducted. Accordingly, the model will
only apply a technology at the particular point deemed suitable. These
constraints are intended to produce results consistent with
manufacturers' technology application practices. For each technology
under consideration, NHTSA stipulates whether it can be applied any
time, at refresh/redesign, or only at redesign. The data forms another
input to the Volpe model. NHTSA develops redesign and refresh schedules
for each of a manufacturer's vehicles included in the analysis,
essentially based on the last known redesign year for each vehicle and
projected forward in a 5-year redesign and a 2-3 year refresh cycle,
and this data is also stored in the market forecast file. We note that
this approach is different than NHTSA has employed previously for
determining redesign and refresh schedules, where NHTSA included the
redesign and refresh dates in the market forecast file as provided by
manufacturers in confidential product plans. The new approach is
necessary given the nature of the new baseline which as a single year
of data does not contain its own refresh and redesign cycle cues for
future model years, and to ensure the complete transparency of the
agency's analysis. Vehicle redesign/refresh assumptions are discussed
in more detail in Section V of the FRIA and in Chapter 3 of the TSD.
NHTSA received comments from the Center for Biological Diversity
(CBD) and Ferrari regarding redesign cycles. CBD stated that
manufacturers do not necessarily adhere to the agencies' assumed five-
year redesign cycle, and may add significant technologies by
redesigning vehicles at more frequent intervals, albeit at higher
costs. CBD argued that NHTSA should analyze the costs and benefits of
manufacturers choosing to redesign vehicles more frequently than a 5-
year average. Conversely, Ferrari agreed with the agencies that major
technology changes are introduced at vehicle redesigns, rather than at
vehicle freshenings, stating further that as compared to full-line
manufacturers, small-volume manufacturers in fact may have 7 to 8-year
redesign cycles. In response, NHTSA recognizes that not all
manufacturers follow a precise five-year redesign cycle for every
vehicle they produce,\583\ but continues to believe that the five-year
redesign cycle assumption is a reasonable estimate of how often
manufacturers can make major technological changes for purposes of its
modeling analysis.\584\ NHTSA has considered attempting to quantify the
increased cost impacts of setting standards that rise in stringency so
rapidly that manufacturers are forced to apply ``usual redesign''
technologies at non-redesign intervals, but such an analysis would be
exceedingly complex and is beyond the scope of this rulemaking given
the timeframe and the current condition of the industry. NHTSA
emphatically disagrees that the redesign cycle is a barrier to
increasing penetration of technologies as CBD suggests, but we also
believe that standards so stringent that they would require
manufacturers to abandon redesign cycles entirely would be beyond the
realm of economic practicability and technological feasibility,
particularly in this rulemaking timeframe given lead time and capital
constraints. Manufacturers can and will accomplish much improvement in
fuel economy and GHG reductions while applying technology consistent
with their redesign schedules.
---------------------------------------------------------------------------

\583\ In prior NHTSA rulemakings, the agency was able to account
for shorter redesign cycles on some models (e.g., some sedans), and
longer redesign cycles on others (e.g., cargo vans), but has
standardized the redesign cycle in this analysis using the
transparent baseline.
\584\ In the MY 2011 final rule, NHTSA noted that the CAR report
submitted by the Alliance, prepared by the Center for Automotive
Research and EDF, stated that ``For a given vehicle line, the time
from conception to first production may span two and one-half to
five years,'' but that ``The time from first production
(``Job1'') to the last vehicle off the line (``Balance
Out'') may span from four to five years to eight to ten years or
more, depending on the dynamics of the market segment,'' The CAR
report then stated that ``At the point of final production of the
current vehicle line, a new model with the same badge and similar
characteristics may be ready to take its place, continuing the
cycle, or the old model may be dropped in favor of a different
product.'' See NHTSA-2008-0089-0170.1, Attachment 16, at 8 (393 of
pdf). NHTSA explained that this description, which states that a
vehicle model will be redesigned or dropped after 4-10 years, was
consistent with other characterizations of the redesign and
freshening process, and supported the 5-year redesign and 2-3 year
refresh cycle assumptions used in the MY 2011 final rule. See id.,
at 9 (394 of pdf). Given that the situation faced by the auto
industry today is not so wholly different from that in March 2009,
when the MY 2011 final rule was published, and given that the
commenters did not present information to suggest that these
assumptions are unreasonable (but rather simply that different
manufacturers may redesign their vehicles more or less frequently,
as the range of cycles above indicates), NHTSA believes that the
assumptions remain reasonable for purposes of this final rule
analysis. See also ``Car Wars 2009-2012, The U.S. automotive product
pipeline,'' John Murphy, Research Analyst, Merrill Lynch research
paper, May 14, 2008 and ``Car Wars 2010-2013, The U.S. automotive
product pipeline,'' John Murphy, Research Analyst, Bank of America/
Merrill Lynch research paper, July 15, 2009. Available at http://
www.autonews.com/assets/PDF/CA66116716.PDF (last accessed March 15,
2010).
---------------------------------------------------------------------------

Once the model indicates that a technology should be applied to a
vehicle, the model must evaluate which technology should be applied.
This will depend on the vehicle subclass to which the vehicle is
assigned; what

[[Page 25574]]

technologies have already been applied to the vehicle (i.e., where in
the ``decision tree'' the vehicle is); when the technology is first
available (i.e., year of availability); whether the technology is still
available (i.e., ``phase-in caps''); and the costs and effectiveness of
the technologies being considered. Technology costs may be reduced, in
turn, by learning effects, while technology effectiveness may be
increased or reduced by synergistic effects between technologies. In
the technology input file, NHTSA has developed a separate set of
technology data variables for each of the twelve vehicle subclasses.
Each set of variables is referred to as an ``input sheet,'' so for
example, the subcompact input sheet holds the technology data that is
appropriate for the subcompact subclass. Each input sheet contains a
list of technologies available for members of the particular vehicle
subclass. The following items are provided for each technology: The
name of the technology, its abbreviation, the decision tree with which
it is associated, the (first) year in which it is available, the upper
and lower cost and effectiveness (fuel consumption reduction)
estimates, the learning type and rate, the cost basis, its
---------------------------------------------------------------------------
applicability, and the phase-in values.

To which vehicle subclass is the vehicle assigned?

As part of its consideration of technological feasibility, the
agency evaluates whether each technology could be implemented on all
types and sizes of vehicles, and whether some differentiation is
necessary in applying certain technologies to certain types and sizes
of vehicles, and with respect to the cost incurred and fuel consumption
and CO2 emissions reduction achieved when doing so. The 2002
NAS Report differentiated technology application using ten vehicle
``classes'' (4 car classes and 6 truck classes),\585\ but did not
determine how cost and effectiveness values differ from class to class.
NAS's purpose in separating vehicles into these classes was to create
groups of ``like'' vehicles, i.e., vehicles similar in size, powertrain
configuration, weight, and consumer use, and for which similar
technologies are applicable. NHTSA similarly differentiates vehicles by
``subclass'' for the purpose of applying technologies to ``like''
vehicles and assessing their incremental costs and effectiveness. NHTSA
assigns each vehicle manufactured in the rulemaking period to one of 12
subclasses: For passenger cars, Subcompact, Subcompact Performance,
Compact, Compact Performance, Midsize, Midsize Performance, Large, and
Large Performance; and for light trucks, Small SUV/Pickup/Van, Midsize
SUV/Pickup/Van, Large SUV/Pickup/Van, and Minivan.
---------------------------------------------------------------------------

\585\ The NAS classes included subcompact cars, compact cars,
midsize cars, large cars, small SUVs, midsize SUVs, large SUVs,
small pickups, large pickups, and minivans.
---------------------------------------------------------------------------

For this final rule as for the NPRM, NHTSA divides the vehicle
fleet into subclasses based on model inputs, and applies subclass-
specific estimates, also from model inputs, of the applicability, cost,
and effectiveness of each fuel-saving technology. Therefore, the
model's estimates of the cost to improve the fuel economy of each
vehicle model depend upon the subclass to which the vehicle model is
assigned.
Each vehicle's subclass is stored in the market forecast file. When
conducting a compliance analysis, if the Volpe model seeks to apply
technology to a particular vehicle, it checks the market forecast to
see if the technology is available and if the refresh/redesign criteria
are met. If these conditions are satisfied, the model determines the
vehicle's subclass from the market data file, which it then uses to
reference another input called the technology input file. NHTSA
reviewed its methodology for dividing vehicles into subclasses for
purposes of technology application that it used in the MY 2011 final
rule, and concluded that the same methodology would be appropriate for
this final rule for MYs 2012-2016. No comments were received on the
vehicle subclasses employed in the agency's NPRM analysis, and NHTSA
has retained the subclasses and the methodology for dividing vehicles
among them for the final rule analysis. Vehicle subclasses are
discussed in more detail in Section V of the FRIA and in Chapter 3 of
the TSD.
For the reader's reference, the subclasses and example vehicles
from the market forecast file are provided in the tables below.

Passenger Car Subclasses Example (MY 2008) Vehicles
------------------------------------------------------------------------
Class Example vehicles
------------------------------------------------------------------------
Subcompact........................ Chevy Aveo, Hyundai Accent.
Subcompact Performance............ Mazda MX-5, BMW Z4.
Compact........................... Chevy Cobalt, Nissan Sentra and
Altima.
Compact Performance............... Audi S4, Mazda RX-8.
Midsize........................... Chevy Impala, Toyota Camry, Honda
Accord, Hyundai Azera.
Midsize Performance............... Chevy Corvette, Ford Mustang (V8),
Nissan G37 Coupe.
Large............................. Audi A8, Cadillac CTS and DTS.
Large Performance................. Bentley Arnage, Daimler CL600.
------------------------------------------------------------------------

Light Truck Subclasses Example (MY 2008) Vehicles
------------------------------------------------------------------------
Class Example vehicles
------------------------------------------------------------------------
Minivans.......................... Dodge Caravan, Toyota Sienna.
Small SUV/Pickup/Van.............. Ford Escape & Ranger, Nissan Rogue.
Midsize SUV/Pickup/Van............ Chevy Colorado, Jeep Wrangler,
Toyota Tacoma.
Large SUV/Pickup/Van.............. Chevy Silverado, Ford E-Series,
Toyota Sequoia.
------------------------------------------------------------------------

What technologies have already been applied to the vehicle (i.e., where
in the ``decision trees'' is it)?

NHTSA's methodology for technology application analysis developed
out of the approach taken by NAS in the 2002 Report, and evaluates the
application of individual technologies and their incremental costs and
effectiveness.

[[Page 25575]]

Incremental costs and effectiveness of individual technologies are
relative to the prior technology state, which means that it is crucial
to understand what technologies are already present on a vehicle in
order to determine correct incremental cost and effectiveness values.
The benefit of the incremental approach is transparency in accounting,
insofar as when individual technologies are added incrementally to
individual vehicles, it is clear and easy to determine how costs and
effectiveness add up as technology levels increase.
To keep track of incremental costs and effectiveness and to know
which technology to apply and in which order, the Volpe model's
architecture uses a logical sequence, which NHTSA refers to as
``decision trees,'' for applying fuel economy-improving technologies to
individual vehicles. In the MY 2011 final rule, NHTSA worked with
Ricardo to modify previously-employed decision trees in order to allow
for a much more accurate application of technologies to vehicles. For
purposes of the final rule, NHTSA reviewed the technology sequencing
architecture and updated, as appropriate, the decision trees used in
the analysis reported in the final rule for MY 2011 and in the MY 2012-
2016 NPRM.
In general, and as described in great detail in the MY 2011 final
rule and in Section V of the current FRIA, each technology is assigned
to one of the five following categories based on the system it affects
or impacts: engine, transmission, electrification/accessory, hybrid or
vehicle. Each of these categories has its own decision tree that the
Volpe model uses to apply technologies sequentially during the
compliance analysis. The decision trees were designed and configured to
allow the Volpe model to apply technologies in a cost-effective,
logical order that also considers ease of implementation. For example,
software or control logic changes are implemented before replacing a
component or system with a completely redesigned one, which is
typically a much more expensive option. In some cases, and as
appropriate, the model may combine the sequential technologies shown on
a decision tree and apply them simultaneously, effectively developing
dynamic technology packages on an as-needed basis. For example, if
compliance demands indicate, the model may elect to apply LUB, EFR, and
ICP on a dual overhead cam engine, if they are not already present, in
one single step. An example simplified decision tree for engine
technologies is provided below; the other simplified decision trees may
be found in Chapter 3 of the Joint TSD and in the FRIA. Expanded
decision trees are available in the docket for this final rule.
BILLING CODE 6560-50-P

[[Page 25576]]

[GRAPHIC] [TIFF OMITTED] TR07MY10.027

BILLING CODE 6560-50-C

[[Page 25577]]

Each technology within the decision trees has an incremental cost
and an incremental effectiveness estimate associated with it, and
estimates are specific to a particular vehicle subclass (see the tables
in Section V of the FRIA). Each technology's incremental estimate takes
into account its position in the decision tree path. If a technology is
located further down the decision tree, the estimates for the costs and
effectiveness values attributed to that technology are influenced by
the incremental estimates of costs and effectiveness values for prior
technology applications. In essence, this approach accounts for ``in-
path'' effectiveness synergies, as well as cost effects that occur
between the technologies in the same path. When comparing cost and
effectiveness estimates from various sources and those provided by
commenters in this and the previous CAFE rulemakings, it is important
that the estimates evaluated are analyzed in the proper context,
especially as concerns their likely position in the decision trees and
other technologies that may be present or missing. Not all estimates
available in the public domain or that have been offered for the
agencies' consideration can be evaluated in an ``apples-to-apples''
comparison with those used by the Volpe model, since in some cases the
order of application, or included technology content, is inconsistent
with that assumed in the decision tree.
The MY 2011 final rule discussed in detail the revisions and
improvements made to the Volpe model and decision trees during that
rulemaking process, including the improved handling and accuracy of
valve train technology application and the development and
implementation of a method for accounting path-dependent correction
factors in order to ensure that technologies are evaluated within the
proper context. The reader should consult the MY 2011 final rule
documents for further information on these modeling techniques, all of
which continued to be utilized in developing this final rule.\586\ To
the extent that the decision trees have changed for purposes of the
NPRM and this final rule, it was due not to revisions in the order of
technology application, but rather to redefinitions of technologies or
addition or subtraction of technologies.
---------------------------------------------------------------------------

\586\ See, e.g., 74 FR 14238-46 (Mar. 30, 2009) for a full
discussion of the decision trees in NHTSA's MY 2011 final rule, and
Docket No. NHTSA-2009-0062-0003.1 for an expanded decision tree used
in that rulemaking.
---------------------------------------------------------------------------

NHTSA did not receive any comments related to the use or ordering
of the decision trees, and the agency continued to use the decision
trees as they were proposed in the NPRM.

Is the next technology available in this model year?

As discussed above, the majority of technologies considered are
available on vehicles today, and thus will be available for application
(albeit in varying degrees) in the model years covered by this rule.
Some technologies, however, will not become available for purposes of
NHTSA's analysis until later in the rulemaking time frame. When the
model is considering whether to add a technology to a vehicle, it
checks its year of availability--if the technology is available, it may
be added; if it is not available, the model will consider whether to
switch to a different decision tree to look for another technology, or
will skip to the next vehicle in a manufacturer's fleet. The year of
availability for each technology is provided above in Table IV.C.2-1.
CBD commented that because many of the technologies considered in
the NPRM are currently available, manufacturers should be able to
attain mpg levels equivalent to the MY 2016 standards in MY 2009. In
response, as discussed above, technology ``availability'' is not
determined based simply on whether the technology exists, but depends
also on whether the technology has achieved a level of technical
viability that makes it appropriate for widespread application. This
depends in turn on component supplier constraints, capital investment
and engineering constraints, and manufacturer product cycles, among
other things. Moreover, even if a technology is available for
application, it may not be available for every vehicle. Some
technologies may have considerable fuel economy benefits, but cannot be
applied to some vehicles due to technological constraints--for example,
cylinder deactivation cannot be applied to vehicles with current 4-
cylinder engines (because not enough cylinders are present to
deactivate some and continue moving the vehicle) or on vehicles with
manual transmissions within the rulemaking timeframe. The agencies have
provided for increases over time to reach the mpg level of the MY 2016
standards precisely because of these types of constraints, because they
have a real effect on how quickly manufacturers can apply technology to
vehicles in their fleets.

Has the technology reached the phase-in cap for this model year?

Besides the refresh/redesign cycles used in the Volpe model, which
constrain the rate of technology application at the vehicle level so as
to ensure a period of stability following any modeled technology
applications, the other constraint on technology application employed
in NHTSA's analysis is ``phase-in caps.'' Unlike vehicle-level cycle
settings, phase-in caps constrain technology application at the vehicle
manufacturer level.\587\ They are intended to reflect a manufacturer's
overall resource capacity available for implementing new technologies
(such as engineering and development personnel and financial
resources), thereby ensuring that resource capacity is accounted for in
the modeling process. At a high level, phase-in caps and refresh/
redesign cycles work in conjunction with one another to avoid the
modeling process out-pacing an OEM's limited pool of available
resources during the rulemaking time frame, especially in years where
many models may be scheduled for refresh or redesign. This helps to
ensure technological feasibility and economic practicability in
determining the stringency of the standards.
---------------------------------------------------------------------------

\587\ While phase-in caps are expressed as specific percentages
of a manufacturer's fleet to which a technology may be applied in a
given model year, phase-in caps cannot always be applied as precise
limits, and the Volpe model in fact allows ``override'' of a cap in
certain circumstances. When only a small portion of a phase-in cap
limit remains, or when the cap is set to a very low value, or when a
manufacturer has a very limited product line, the cap might prevent
the technology from being applied at all since any application would
cause the cap to be exceeded. Therefore, the Volpe model evaluates
and enforces each phase-in cap constraint after it has been exceeded
by the application of the technology (as opposed to evaluating it
before application), which can result in the described overriding of
the cap.
---------------------------------------------------------------------------

NHTSA has been developing the concept of phase-in caps for purposes
of the agency's modeling analysis over the course of the last several
CAFE rulemakings, as discussed in greater detail in the MY 2011 final
rule,\588\ and in Section V of the FRIA and Chapter 3 of the Joint TSD.
The MY 2011 final rule employed non-linear phase-in caps (that is, caps
that varied from year to year) that were designed to respond to
comments raising lead-time concerns in reference to the agency's
proposed MY 2011-2015 standards, but because the final rule covered
only one model year, many phase-in caps for that model year were lower
than had originally been proposed. NHTSA emphasized that the MY 2011
phase-in caps were based on assumptions for the full five year period
of the proposal (2011-2015), and stated that it would reconsider the
phase-in settings for all years beyond 2011 in a future rulemaking
analysis.\589\
---------------------------------------------------------------------------

\588\ 74 FR 14268-14271 (Mar. 30, 2009).
\589\ See 74 FR at 14269 (Mar. 20, 2009).

---------------------------------------------------------------------------

[[Page 25578]]

For purposes of this final rule for MYs 2012-2016, as in the MY
2011 final rule, NHTSA combines phase-in caps for some groups of
similar technologies, such as valve phasing technologies that are
applicable to different forms of engine design (SOHC, DOHC, OHV), since
they are very similar from an engineering and implementation
standpoint. When the phase-in caps for two technologies are combined,
the maximum total application of either or both to any manufacturer's
fleet is limited to the value of the cap.\590\ In contrast to the
phase-in caps used in the MY 2011 final rule, NHTSA has increased the
phase-in caps for most of the technologies, as discussed below.
---------------------------------------------------------------------------

\590\ See 74 FR at 14270 (Mar. 30, 2009) for further discussion
and examples.
---------------------------------------------------------------------------

In developing phase-in cap values for purposes of this final rule,
NHTSA initially considered the fact that many of the technologies
commonly applied by the model, those placed near the top of the
decision trees, such as low friction lubes, valve phasing, electric
power steering, improved automatic transmission controls, and others,
have been commonly available to manufacturers for several years now.
Many technologies, in fact, precede the 2002 NAS Report, which
estimated that such technologies would take 4 to 8 years to penetrate
the fleet. Since this final rule would take effect in MY 2012, nearly
10 years beyond the NAS report, and extends to MY 2016, and in the
interest of harmonization with EPA's proposal, NHTSA determined that
higher phase-in caps were likely justified. Additionally, NHTSA
considered the fact that manufacturers, as part of the agreements
supporting the National Program, appear to be anticipating higher
technology application rates than those used in the MY 2011 final rule.
This also supported higher phase-in caps for purposes of the analysis
underlying this final rule.
Thus, while phase-in caps for the MY 2011 final rule reached a
maximum of 50 percent for a couple of technologies and generally fell
in the range between 0 and 20 percent, phase-in caps for this final
rule for the majority of technologies are set to reach 85 or 100
percent by MY 2016, although more advanced technologies like diesels
and strong hybrids reach only 15 percent by MY 2016.
NHTSA received comments from the Alliance and ICCT relating to
phase-in caps. The Alliance commented that the higher phase-in caps in
the NPRM analysis (as compared to the MY 2011 final rule) ``ignore OEM
engine architecture differences/limitations,'' arguing that the agency
must consider manufacturing investment and lead time implications when
defining phase-in caps. ICCT did not raise the issue of phase-in caps
directly, but commented that the agencies had not provided information
in the proposal documents explaining when each manufacturer can
implement the different technologies and how long it will take the
technologies to spread across the fleet. ICCT argued that this
information was crucial to considering how quickly the stringency of
the standards could be increased, and at what cost.
In response to the Alliance comments, the phase-in cap constraint
is, in fact, exactly intended to account for manufacturing investment
and lead time implications, as discussed above: phase-in caps are
intended to reflect a manufacturer's overall resource capacity
available for implementing new technologies (such as engineering and
development personnel and financial resources), to help ensure that
resource capacity is accounted for in the modeling process. Although
the phase-in caps for the analysis supporting these standards are
higher than the phase-in caps employed in the MY 2011 final rule, as
stated in the NPRM, the agencies considered the fact that
manufacturers, as part of the agreements supporting the National
Program, appear to be anticipating higher technology application rates
during the rulemaking timeframe--indicating that the values selected
for the phase-in caps are more likely within the range of
practicability. Additionally, the agencies did not receive any comments
from manufacturers indicating a direct concern with the proposed
application rates, which they were able to review in the detailed
manufacturer level model outputs. The agencies believe that as
manufacturers focus their resources (i.e., engineering, capital
investment, etc.) on fuel economy-improving technologies, many of which
have been in production for many years, the application rates being
modeled are appropriate for the timeframe being analyzed.
In response to ICCT's comments, the combination of phase-in caps,
refresh/redesign cycles, engineering constraints, etc., are intended to
simulate manufacturers' technology application decisions, and
ultimately define the technology application/implementation rates for
each manufacturer. NHTSA has used the best public data available to
define refresh and redesign schedules to define technology
implementation, which allows us to apply technologies at the specific
times each manufacturer is planning. There was full notice of not just
the phase-in caps themselves, but their specific application as well.
NHTSA notes that the PRIA and the FRIA do contain manufacturer-specific
application/implementation rates for prominent technologies, and that
manufacturer-specific technology application as employed in the
agency's analysis is available in full in the Volpe model outputs
available on NHTSA's Web site. The model outputs present the resultant
application of technologies at the industry, manufacturer, and vehicle
levels.
Theoretically, significantly higher phase-in caps, such as those
used in the current proposal and final rule as compared to those used
in the MY 2011 final rule, should result in higher levels of technology
penetration in the modeling results. Reviewing the modeling output does
not, however, indicate unreasonable levels of technology penetration
for the final standards.\591\ NHTSA believes that this is due to the
interaction of the various changes in methodology for this final rule--
changes to phase-in caps are but one of a number of revisions to the
Volpe model and its inputs that could potentially impact the rate at
which technologies are applied in the modeling analysis for this final
rule as compared to prior rulemakings. Other revisions that could
impact modeled application rates include the use of transparent CAFE
certification data in baseline fleet formulation and the use of other
data for projecting it forward,\592\ or the use of a multi-year
planning programming technique to apply technology retroactively to
earlier-MY vehicles, both of which may have a direct impact on the
modeling process. Conversely the model and inputs remain unchanged in
other areas that also could impact technology application, such as in
the refresh/redesign cycle settings, estimates used for the
technologies, both of which remain largely unchanged from the MY 2011
final rule. These changes together make it difficult to predict how
phase-in caps should be expected to function in the new modeling
process.
---------------------------------------------------------------------------

\591\ The modeling output for the analysis underlying these
final standards is available on NHTSA's Web site.
\592\ The baseline fleet sets the starting point, from a
technology point of view, for where the model begins the technology
application process, so changes have a direct impact on the
projected net application of technology.
---------------------------------------------------------------------------

Thus, after reviewing the output files, NHTSA concludes that the
higher phase-in caps, and the resulting technology application rates
produced by the Volpe model, at both the industry and manufacturer
level, are appropriate for the analysis underlying these final

[[Page 25579]]

standards, achieving a suitable level of stringency without requiring
---------------------------------------------------------------------------
unrealistic or unachievable penetration rates.

Is the technology less expensive due to learning effects?

Historically, NHTSA did not explicitly account for the cost
reductions a manufacturer might realize through learning achieved from
experience in actually applying a technology. Since working with EPA to
develop the 2008 NPRM for MYs 2011-2015, and with Ricardo to refine the
concept for the March 2009 MY 2011 final rule, NHTSA has accounted for
these cost reductions through two kinds of mutually exclusive learning,
``volume-based'' and ``time-based'' which it continues to use in this
rule, as discussed below.
In the 2008 NPRM, NHTSA applied learning factors to technology
costs for the first time. These learning factors were developed using
the parameters of learning threshold, learning rate, and the initial
cost, and were based on the ``experience curve'' concept which
describes reductions in production costs as a function of accumulated
production volume. The typical curve shows a relatively steep initial
decline in cost which flattens out to a gentle downwardly sloping line
as the volume increase to large values. In the NPRM, NHTSA applied a
learning rate discount of 20 percent for each successive doubling of
production volume (on a per manufacturer basis), and a learning
threshold of 25,000 units was assumed (thus a technology was viewed as
being fully learned out at 100,000 units). The factor was only applied
to certain technologies that were considered emerging or newly
implemented on the basis that significant cost improvements would be
achieved as economies of scale were realized (i.e., the technologies
were on the steep part of the curve).
In the MY 2011 final rule, NHTSA continued to use this learning
factor, referring to it as volume-based learning since the cost
reductions were determined by production volume increases, and again
only applied it to emerging technologies. However, and in response to
comments, NHTSA revised its assumptions on learning threshold, basing
them instead on an industry-wide production basis, and increasing the
threshold to 300,000 units annually.
Commenters to the 2008 NPRM also described another type of learning
factor which NHTSA decided to adopt and implement in the MY 2011 final
rule. Commenters described a relatively small negotiated cost decrease
that occurred on an annual basis through contractual agreements with
first tier component and systems suppliers for readily available, high
volume technologies commonly in use by multiple OEMs. Based on the same
experience curve principal, however at production volumes that were on
the flatter part of the curve (and thus the types of volumes that
represent annual industry volumes), NHTSA adopted this type learning
and referred to it as time-based learning. An annual cost reduction of
3 percent in the second and each subsequent year, which was consistent
with estimates from commenters and supported by work Ricardo conducted
for NHTSA, was used in the final rule.
In developing the proposed standards, NHTSA and EPA reviewed both
types of learning factors, and the thresholds (300,000) and reduction
rates (20 percent for volume, 3 percent for time-based) they rely on,
and as implemented in the MY 2011 final rule, and agreed that both
factors continue to be accurate and appropriate; each agency thus
implemented time- and volume-based learning in their analyses. Noting
that only one type of learning can be applied to any single technology,
if any learning is applied at all, the agencies reviewed each to
determine which learning factor was appropriate. Volume-based learning
was applied to the higher complexity hybrid technologies, while no
learning was applied to technologies likely to be affected by commodity
costs (LUB, ROLL) or that have loosely-defined BOMs (EFR, LDB), as was
the case in the MY 2011 final rule. Chapter 3 of the Joint TSD shows
the specific learning factors that NHTSA has applied in this analysis
for each technology, and discusses learning factors and each agencies'
use of them further.
ICCT and Ferrari commented on learning curves. ICCT stated the
agencies could improve the accuracy of the learning curve assumptions
if they used a more dynamic or continuous learning curve that is more
technology-specific, rather than using step decreases as the current
time- and volume-based learning curves appear to do. ICCT also
commented on the appropriate application of volume- versus time-based
learning, and stated further that worldwide production volumes should
be taken into account when developing learning curves. Ferrari
commented that is more difficult for small-volume manufacturers to
negotiate cost decreases from things like cost learning effects with
their suppliers, implying that learning effects may not be applicable
equally for all manufacturers.
NHTSA agrees that a continuous curve, if implemented correctly,
could potentially improve the accuracy of modeling cost-learning
effects, although the agency cannot estimate at this time how
significant the improvement would be. To implement a continuous curve,
however, NHTSA would need to develop a learning curve cost model to be
integrated into the agency's existing model for CAFE analysis. Due to
time constraints the agencies were not able to investigate fully the
use of a continuous cost-learning effects curve for each technology,
but we will investigate the applicability of this approach for future
rulemakings. For purposes of the final rule analysis, however, NHTSA
believes that while more detailed cost learning approaches may
eventually be possible, the approach taken for this final rule is
valid.
Additionally, while the agencies agree that worldwide production
volumes can impact learning curves, the agencies do not forecast
worldwide vehicle production volumes in addition to the already complex
task of forecasting the U.S. market. That said, the agencies do
consider current and projected worldwide technology proliferation when
determining the maturity of a particular technology used to determine
the appropriateness of applying time- or volume-based learning, which
helps to account for the effect of globalized production.
With regard to ICCT's comments on the appropriate application of
volume- versus time-based learning, however, it seems as though ICCT is
referencing a study that defines volume- and time-based learning in a
different manner than the current definitions used by the agencies, and
so is not directly relevant. The agencies use ``volume-based'' learning
for non-mature technologies that have the potential for significant
cost reductions through learning, while ``time-based'' learning is used
for mature technologies that have already had significant cost
reductions and only have the potential for smaller cost reductions. For
``time-based'' learning, the agencies chose to emulate the small year-
over-year cost reductions manufacturers realize through defined cost
reductions, approximately 3 percent per year, negotiated into contracts
with suppliers. A more detailed description of how the agencies define
volume- and time-based learning can be found in NHTSA's PRIA.
And finally, in response to Ferrari's comment, NHTSA recognizes
that cost negotiations can be different for different manufacturers,
but believes that on balance, cost learning at the supplier level will
generally impact costs to all purchasers. Thus, if cost reductions are
realized for a particular

[[Page 25580]]

technology, all entities that purchase the technology will benefit from
these cost reductions.
Is the technology more or less effective due to synergistic effects?
When two or more technologies are added to a particular vehicle
model to improve its fuel efficiency and reduce CO2
emissions, the resultant fuel consumption reduction may sometimes be
higher or lower than the product of the individual effectiveness values
for those items.\593\ This may occur because one or more technologies
applied to the same vehicle partially address the same source (or
sources) of engine, drivetrain or vehicle losses. Alternately, this
effect may be seen when one technology shifts the engine operating
points, and therefore increases or reduces the fuel consumption
reduction achieved by another technology or set of technologies. The
difference between the observed fuel consumption reduction associated
with a set of technologies and the product of the individual
effectiveness values in that set is referred to for purposes of this
rulemaking as a ``synergy.'' Synergies may be positive (increased fuel
consumption reduction compared to the product of the individual
effects) or negative (decreased fuel consumption reduction). An example
of a positive synergy might be a vehicle technology that reduces road
loads at highway speeds (e.g., lower aerodynamic drag or low rolling
resistance tires), that could extend the vehicle operating range over
which cylinder deactivation may be employed. An example of a negative
synergy might be a variable valvetrain system technology, which reduces
pumping losses by altering the profile of the engine speed/load map,
and a six-speed automatic transmission, which shifts the engine
operating points to a portion of the engine speed/load map where
pumping losses are less significant. As the complexity of the
technology combinations is increased, and the number of interacting
technologies grows accordingly, it becomes increasingly important to
account for these synergies.
---------------------------------------------------------------------------

\593\ More specifically, the products of the differences between
one and the technology-specific levels of effectiveness in reducing
fuel consumption. For example, not accounting for interactions, if
technologies A and B are estimated to reduce fuel consumption by 10
percent (i.e., 0.1) and 20 percent (i.e., 0.2) respectively, the
``product of the individual effectiveness values'' would be 1-0.1
times 1-0.2, or 0.9 times 0.8, which equals 0.72, corresponding to a
combined effectiveness of 28 percent rather than the 30 percent
obtained by adding 10 percent to 20 percent. The ``synergy factors''
discussed in this section further adjust these multiplicatively
combined effectiveness values.
---------------------------------------------------------------------------

NHTSA and EPA determined synergistic impacts for this rulemaking
using EPA's ``lumped parameter'' analysis tool, which EPA described at
length in its March 2008 Staff Technical Report.\594\ The lumped
parameter tool is a spreadsheet model that represents energy
consumption in terms of average performance over the fuel economy test
procedure, rather than explicitly analyzing specific drive cycles. The
tool begins with an apportionment of fuel consumption across several
loss mechanisms and accounts for the average extent to which different
technologies affect these loss mechanisms using estimates of engine,
drivetrain and vehicle characteristics that are averaged over the EPA
fuel economy drive cycle. Results of this analysis were generally
consistent with those of full-scale vehicle simulation modeling
performed in 2007 by Ricardo, Inc.
---------------------------------------------------------------------------

\594\ EPA Staff Technical Report: Cost and Effectiveness
Estimates of Technologies Used to Reduce Light-duty Vehicle Carbon
Dioxide Emissions; EPA420-R-08-008, March 2008. Available at Docket
No. NHTSA-2009-0059-0027.
---------------------------------------------------------------------------

For the current rulemaking, NHTSA used the lumped parameter tool as
modified in the MY 2011 CAFE final rule. NHTSA modified the lumped
parameter tool from the version described in the EPA Staff Technical
Report in response to public comments received in that rulemaking. The
modifications included updating the list of technologies and their
associated effectiveness values to match the updated list of
technologies used in the final rule. NHTSA also expanded the list of
synergy pairings based on further consideration of the technologies for
which a competition for losses would be expected. These losses are
described in more detail in Section V of the FRIA.
NHTSA and EPA incorporate synergistic impacts in their analyses in
slightly different manners. Because NHTSA applies technologies
individually in its modeling analysis, NHTSA incorporates synergistic
effects between pairings of individual technologies. The use of
discrete technology pair incremental synergies is similar to that in
DOE's National Energy Modeling System (NEMS).\595\ Inputs to the Volpe
model incorporate NEMS-identified pairs, as well as additional pairs
from the set of technologies considered in the Volpe model.
---------------------------------------------------------------------------

\595\ U.S. Department of Energy, Energy Information
Administration, Transportation Sector Module of the National Energy
Modeling System: Model Documentation 2007, May 2007, Washington, DC,
DOE/EIAM070(2007), at 29-30. Available at http://tonto.eia.doe.gov/
ftproot/modeldoc/m070(2007).pdf (last accessed March 15, 2010).
---------------------------------------------------------------------------

NHTSA notes that synergies that occur within a decision tree are
already addressed within the incremental values assigned and therefore
do not require a synergy pair to address. For example, all engine
technologies take into account incremental synergy factors of preceding
engine technologies, and all transmission technologies take into
account incremental synergy factors of preceding transmission
technologies. These factors are expressed in the fuel consumption
improvement factors in the input files used by the Volpe model.
For applying incremental synergy factors in separate path
technologies, the Volpe model uses an input table (see the tables in
Chapter 3 of the TSD and in the FRIA) which lists technology pairings
and incremental synergy factors associated with those pairings, most of
which are between engine technologies and transmission/electrification/
hybrid technologies. When a technology is applied to a vehicle by the
Volpe model, all instances of that technology in the incremental
synergy table which match technologies already applied to the vehicle
(either pre-existing or previously applied by the Volpe model) are
summed and applied to the fuel consumption improvement factor of the
technology being applied. Synergies for the strong hybrid technology
fuel consumption reductions are included in the incremental value for
the specific hybrid technology block since the model applies
technologies in the order of the most effectiveness for least cost and
also applies all available electrification and transmission
technologies before applying strong hybrid technologies.
NHTSA received only one comment regarding synergies, from MEMA, who
commented that NHTSA's Volpe model adequately addressed synergistic
effects. Having received no information to the contrary, NHTSA
finalized the synergy approach and values for the final rule.
d. Where can readers find more detailed information about NHTSA's
technology analysis?
Much more detailed information is provided in Section V of the
FRIA, and a discussion of how NHTSA and EPA jointly reviewed and
updated technology assumptions for purposes of this final rule is
available in Chapter 3 of the TSD. Additionally, all of NHTSA's model
input and output files are now public and available for the reader's
review and consideration. The technology input files can be found in
the docket for this final rule, Docket No. NHTSA-2009-0059, and on
NHTSA's

[[Page 25581]]

Web site. And finally, because much of NHTSA's technology analysis for
purposes of this final rule builds on the work that was done for the MY
2011 final rule, we refer readers to that document as well for
background information concerning how NHTSA's methodology for
technology application analysis has evolved over the past several
rulemakings, both in response to comments and as a result of the
agency's growing experience with this type of analysis.\596\
---------------------------------------------------------------------------

\596\ 74 FR 14233-308 (Mar. 30, 2009).
\597\ The $21 value is for CO2 emissions in 2010, which rises to
$45/ton in 2050, at an average discount rate of 3 percent.
---------------------------------------------------------------------------

3. How did NHTSA develop its economic assumptions?
NHTSA's analysis of alternative CAFE standards for the model years
covered by this rulemaking relies on a range of forecast variables,
economic assumptions, and parameter values. This section describes the
sources of these forecasts, the rationale underlying each assumption,
and the agency's choices of specific parameter values. These economic
values play a significant role in determining the benefits of
alternative CAFE standards, as they have for the last several CAFE
rulemakings. Under those alternatives where standards would be
established by reference to their costs and benefits, these economic
values also affect the levels of the CAFE standards themselves. Some of
these variables have more important effects on the level of CAFE
standards and the benefits from requiring alternative increases in fuel
economy than do others.
In reviewing these variables and the agency's estimates of their
values for purposes of this final rule, NHTSA reconsidered previous
comments it had received and comments received to the NPRM, as well as
reviewed newly available literature. As a consequence, the agency
elected to revise some of its economic assumptions and parameter
estimates from previous rulemakings at the NPRM stage, while retaining
others. Some of the most important changes, which are discussed in
greater detail below, as well as in Chapter 4 of the Joint TSD and in
Chapter VIII of the FRIA, include significant revisions to the markup
factors for technology costs; reducing the rebound effect from 15 to 10
percent; and revising the value of reducing CO2 emissions
based on recent interagency efforts to develop estimates of this value
for government-wide use. The comments the agency received and its
responses are discussed in detail below, as well as in the TSD and
FRIA. For the reader's reference, Table IV.C.3-1 below summarizes the
values used to calculate the economic benefits from each alternative.

Table IV.C.3-1--Economic Values for Benefits Computations
[2007$]
------------------------------------------------------------------------

------------------------------------------------------------------------
Fuel Economy Rebound Effect............. 10%
``Gap'' between test and on-road MPG.... 20%
Value of refueling time per ($ per $24.64
vehicle-hour).
Average percentage of tank refilled per 55%
refueling.
Percent of drivers refueling in response 100%
to low fuel level.
Annual growth in average vehicle use.... 1.15%
Fuel Prices (2012-50 average, $/gallon) ..............................
Retail gasoline price............... $3.66
Pre-tax gasoline price.............. $3.29
Economic Benefits from Reducing Oil ..............................
Imports ($/gallon)
``Monopsony'' Component............. $0.00
Price Shock Component............... $0.17
Military Security Component......... $0.00
-------------------------------
Total Economic Costs ($/gallon). $0.17
Emission Damage Costs (2020, $/ton or $/ ..............................
metric ton)
Carbon monoxide..................... $0
Volatile organic compounds (VOC).... $1,300
Nitrogen oxides (NOx)--vehicle use.. $5,300
Nitrogen oxides (NOx)--fuel $5,100
production and distribution.
Particulate matter (PM2.5)--vehicle $290,000
use.
Particulate matter (PM2.5)--fuel $240,000
production and distribution.
Sulfur dioxide (SO2)................ $31,000
Carbon dioxide (CO2)................ $21 \597\
Annual Increase in CO2 Damage Cost...... Varies by year.
External Costs from Additional ..............................
Automobile Use ($/vehicle-mile)
Congestion.......................... $0.054
Accidents........................... $0.023
Noise............................... $0.001
-------------------------------
Total External Costs............ $0.078
External Costs from Additional Light ..............................
Truck Use ($/vehicle-mile)
Congestion.......................... $0.048
Accidents........................... $0.026
Noise............................... $0.001
-------------------------------
Total External Costs............ $0.075
Discount Rate Applied to Future Benefits 3%, 7%
------------------------------------------------------------------------

[[Page 25582]]

a. Costs of Fuel Economy-Improving Technologies
NHTSA and EPA previously developed detailed estimates of the costs
of applying fuel economy-improving technologies to vehicle models for
use in analyzing the impacts of alternative standards considered in the
proposed rulemaking, including varying cost estimates for applying
certain fuel economy technologies to vehicles of different sizes and
body styles. These estimates were modified for purposes of this
analysis as a result of extensive consultations among engineers from
NHTSA, EPA, and the Volpe Center. Building on NHTSA's estimates
developed for the MY 2011 CAFE final rule and EPA's Advanced Notice of
Proposed Rulemaking, which relied on EPA's 2008 Staff Technical Report,
the two agencies took a fresh look at technology cost and effectiveness
values and incorporated FEV tear-down study results for purposes of
this joint final rule under the National Program.
While NHTSA generally found that much of the cost information used
in the MY 2011 final rule and EPA's 2008 Staff Report was consistent to
a great extent, the agencies, in reconsidering information from many
sources, revised the component costs of several major technologies
including: turbocharging/downsizing, mild and strong hybrids, diesels,
SGDI, and Valve Train Lift Technologies for purposes of the NPRM. In
addition, based on FEV tear-down studies, the costs for turbocharging/
downsizing, 6-, 7-, 8-speed automatic transmissions, and dual clutch
transmissions were revised for this final rule.
The technology cost estimates used in this analysis are intended to
represent manufacturers' direct costs for high-volume production of
vehicles with these technologies and sufficient experience with their
application so that all remaining cost reductions due to ``learning
curve'' effects have been fully realized. However, NHTSA recognizes
that manufacturers' actual costs for employing these technologies
include additional outlays for accompanying design or engineering
changes to models that use them, development and testing of prototype
versions, recalibrating engine operating parameters, and integrating
the technology with other attributes of the vehicle. Manufacturers'
indirect costs for employing these technologies also include expenses
for product development and integration, modifying assembly processes
and training assembly workers to install them, increased expenses for
operation and maintaining assembly lines, higher initial warranty costs
for new technologies, any added expenses for selling and distributing
vehicles that use these technologies, and manufacturer and dealer
profit.
In previous CAFE rulemakings and in NHTSA's safety rulemakings, the
agency has accounted for these additional costs by using a Retail Price
Equivalent (RPE) multiplier of 1.5. For purposes of this rulemaking,
based on recent work by EPA, NHTSA has applied indirect cost
multipliers ranging from 1.11 to 1.64 to the estimates of vehicle
manufacturers' direct costs for producing or acquiring each technology
to improve fuel economy.\598\ These multipliers vary with the
complexity of each technology and the time frame over which costs are
estimated. More complex technologies are associated with higher
multipliers because of the larger increases in manufacturers' indirect
costs for developing, producing (or procuring), and deploying these
more complex technologies. The appropriate multipliers decline over
time for technologies of all complexity levels, since increased
familiarity and experience with their application is assumed to reduce
manufacturers' indirect costs for employing them.
---------------------------------------------------------------------------

\598\ NHTSA notes that in addition to the technology cost
analysis employing this ``ICM'' approach, the FRIA contains a
sensitivity analysis using a technology cost multiplier of 1.5.
---------------------------------------------------------------------------

NHTSA and EPA received far fewer specific comments on technology
cost estimates than in previous CAFE rulemakings, which suggests that
most, although not all, stakeholders generally agreed with the
agencies' assumptions. Several commenters supported the agencies' use
of tear-down studies for developing some of the technology costs,
largely citing the agencies' own reasons in support of that
methodology. Some specific comments were received with regard to hybrid
and other technology costs, to which the agencies are responding
directly in Chapter 3 of the Joint TSD and in the agencies' respective
FRIAs. Generally speaking, however, to the extent that commenters
disagreed with the agencies' cost estimates, often the disagreement
stemmed from assumptions about the technology's maturity, which the
agencies have tried to account for in the analysis. These issues are
discussed further in Chapter 3 of the TSD. Additionally, we note that
technology costs will also be addressed in the upcoming revised NAS
report.
With regard to the indirect cost multiplier approach, commenters
also generally supported the higher level of specificity provided by
the ICM approach compared to the RPE approach, although some commenters
suggested specific refinements to the measurement of ICMs. For example,
while the automotive dealer organization NADA argued that all dealer
costs of sales should be included in ``dealer profit,'' another
commenter noted expressly that the ICM does not include profits.
Comments from ICCT also argued in favor of revising the ``technology
complexity'' component of the ICM to account for the complexity of
integrating a new technology into a vehicle, rather than for only the
complexity of producing the technology itself. These comments and
others on the ICM are addressed in Chapter 3 of the Joint TSD and in
the agencies' respective FRIAs. NHTSA notes that profits were not
included in the indirect cost estimates of this rule, and also that
NHTSA's sensitivity analysis, presented in Chapter X of the FRIA,
indicates that using the 1.5 RPE multiplier would result in higher
costs compared to today's final rule costs incorporating the ICM
multiplier, although even with those higher costs the 1.5 RPE analysis
still resulted in significant net benefits for the rulemaking as a
whole. NHTSA continues to study this issue and may employ a different
approach in future rulemakings.
b. Potential Opportunity Costs of Improved Fuel Economy
An important concern is whether achieving the fuel economy
improvements required by alternative CAFE standards might result in
manufacturers compromising the performance, carrying capacity, safety,
or comfort of their vehicle models. To the extent that it does so, the
resulting sacrifice in the value of these attributes to consumers
represents an additional cost of achieving the required improvements in
fuel economy. (This possibility is addressed in detail in Section
IV.G.6.) Although exact dollar values of these attributes to consumers
are difficult to infer, differences in vehicle purchase prices and
buyers' choices among competing models that feature varying
combinations of these characteristics clearly demonstrate that changes
in these attributes affect the utility and economic value that vehicles
offer to potential buyers.\599\
---------------------------------------------------------------------------

\599\ See, e.g., Kleit A.N., 1990. ``The Effect of Annual
Changes in Automobile Fuel Economy Standards.'' Journal of
Regulatory Economics 2: 151-172 (Docket EPA-HQ-OAR-2009-0472-0015);
Berry, Steven, James Levinsohn, and Ariel Pakes, 1995. ``Automobile
Prices in Market Equilibrium,'' Econometrica 63(4): 841-940 (Docket
NHTSA-2009-0059-0031); McCarthy, Patrick S., 1996. ``Market Price
and Income Elasticities of New Vehicle Demands.'' Review of
Economics and Statistics 78: 543-547 (Docket NHTSA-2009-0059-0039);
and Goldberg, Pinelopi K., 1998. ``The Effects of the Corporate
Average Fuel Efficiency Standards in the U.S.,'' Journal of
Industrial Economics 46(1): 1-33 (Docket EPA-HQ-OAR-2009-0472-0017).

---------------------------------------------------------------------------

[[Page 25583]]

NHTSA and EPA have approached this potential problem by developing
cost estimates for fuel economy-improving technologies that include any
additional manufacturing costs that would be necessary to maintain the
originally planned levels of performance, comfort, carrying capacity,
and safety of any light-duty vehicle model to which those technologies
are applied. In doing so, the agencies followed the precedent
established by the 2002 NAS Report, which estimated ``constant
performance and utility'' costs for fuel economy technologies. NHTSA
has used these as the basis for its continuing efforts to refine the
technology costs it uses to analyze manufacturer's costs for complying
with alternative passenger car and light truck CAFE standards for MYs
2012-2016. Although the agency has revised its estimates of
manufacturers' costs for some technologies significantly for use in
this rulemaking, these revised estimates are still intended to
represent costs that would allow manufacturers to maintain the
performance, carrying capacity, and utility of vehicle models while
improving their fuel economy.
Although we believe that our cost estimates for fuel economy-
improving technologies include adequate provision for accompanying
outlays that are necessary to prevent any significant degradation in
other attributes that vehicle owners value, it is possible that they do
not include adequate allowance for the necessary efforts by
manufacturers to prevent sacrifices in these attributes on all vehicle
models. If this is the case, the true economic costs of achieving
higher fuel economy should include the opportunity costs to vehicle
owners of any sacrifices in vehicles' performance, carrying capacity,
and utility, and omitting these will cause the agency's estimated
technology costs to underestimate the true economic costs of improving
fuel economy.
Recognizing this possibility, it would be desirable to estimate
explicitly the changes in vehicle buyers' welfare from the combination
of higher prices for new vehicle models, increases in their fuel
economy, and any accompanying changes in vehicle attributes such as
performance, passenger- and cargo-carrying capacity, or other
dimensions of utility. The net change in buyer's welfare that results
from the combination of these changes would provide a more accurate
estimate of the true economic costs for improving fuel economy.
Although the agency has been unable to develop a procedure for doing so
as part of this rulemaking, Section IV.G.6. below includes a detailed
analysis and discussion of how omitting possible changes in vehicle
attributes other than their prices and fuel economy might affect its
estimates of benefits and costs resulting from the standards this rule
establishes.
c. The On-Road Fuel Economy ``Gap''
Actual fuel economy levels achieved by light-duty vehicles in on-
road driving fall somewhat short of their levels measured under the
laboratory-like test conditions used by EPA to establish its published
fuel economy ratings for different models. In analyzing the fuel
savings from alternative CAFE standards, NHTSA has previously adjusted
the actual fuel economy performance of each light truck model downward
from its rated value to reflect the expected size of this on-road fuel
economy ``gap.'' On December 27, 2006, EPA adopted changes to its
regulations on fuel economy labeling, which were intended to bring
vehicles' rated fuel economy levels closer to their actual on-road fuel
economy levels.\600\
---------------------------------------------------------------------------

\600\ 71 FR 77871 (Dec. 27, 2006).
---------------------------------------------------------------------------

In its Final Rule, EPA estimated that actual on-road fuel economy
for light-duty vehicles averages 20 percent lower than published fuel
economy levels. For example, if the overall EPA fuel economy rating of
a light truck is 20 mpg, the on-road fuel economy actually achieved by
a typical driver of that vehicle is expected to be 16 mpg (20*.80).
NHTSA employed EPA's revised estimate of this on-road fuel economy gap
in its analysis of the fuel savings resulting from alternative CAFE
standards evaluated in the MY 2011 final rule.
For purposes of this final rule, NHTSA conducted additional
analysis of this issue. The agency used data on the number of passenger
cars and light trucks of each model year that were registered for use
during calendar years 2000 through 2006, average rated fuel economy for
passenger cars and light trucks produced during each model year, and
estimates of average miles driven per year by cars and light trucks of
different ages. These data were combined to develop estimates of the
average fuel economy that the U.S. passenger vehicle fleet would have
achieved from 2000 through 2006 if cars and light trucks of each model
year achieved the same fuel economy levels in actual on-road driving as
they did under test conditions when new.
NHTSA compared these estimates to the Federal Highway
Administration's (FHWA) published values of actual on-road fuel economy
for passenger cars and light trucks during each of those years.\601\
FHWA's estimates of actual fuel economy for passenger cars averaged 22
percent lower than NHTSA's estimates of its fleet-wide average value
under test conditions over this period, while FHWA's estimates for
light trucks averaged 17 lower than NHTSA's estimates of average light
truck fuel economy under test conditions. These results appear to
confirm that the 20 percent on-road fuel economy discount or gap
represents a reasonable estimate for use in evaluating the fuel savings
likely to result from alternative CAFE standards for MY 2012-2016
vehicles.
---------------------------------------------------------------------------

\601\ Federal Highway Administration, Highway Statistics, 2000
through 2006 editions, Table VM-1; See http://www.fhwa.dot.gov/
policy/ohpi/hss/hsspubs.cfm (last accessed March 1, 2010).
---------------------------------------------------------------------------

NHTSA received no comments on this issue in response to the NPRM.
Accordingly, it has not revised its estimate of the on-road fuel
economy gap from the 20 percent figure used previously.
d. Fuel Prices and the Value of Saving Fuel
Projected future fuel prices are a critical input into the economic
analysis of alternative CAFE standards, because they determine the
value of fuel savings both to new vehicle buyers and to society. NHTSA
relied on the most recent fuel price projections from the U.S. Energy
Information Administration's (EIA) Annual Energy Outlook (AEO) for this
analysis. Specifically, we used the AEO 2010 Early Release (December
2009) Reference Case forecasts of inflation-adjusted (constant-dollar)
retail gasoline and diesel fuel prices, which represent the EIA's most
up-to-date estimate of the most likely course of future prices for
petroleum products.\602\ This forecast is

[[Page 25584]]

somewhat lower than the AEO 2009 Reference Case forecast the agency
relied upon in the analysis it conducted for the NPRM. Over the period
from 2010 to 2030, the AEO 2010 Early Release Reference Case forecast
of retail gasoline prices used in this analysis averages $3.18 per
gallon (in 2007 dollars), in contrast to the $3.38 per gallon average
price for that same period forecast in the earlier AEO 2009 Reference
Case and used in the NPRM analysis.
---------------------------------------------------------------------------

\602\ Energy Information Administration, Annual Energy Outlook
2010 Early Release, Reference Case (December 2009), Table A12.
Available at http://www.eia.doe.gov/oiaf/aeo/pdf/appa.pdf, p. 25
(last accessed March 1, 2010). These forecasts reflect the
provisions of the Energy Independence and Security Act of 2007
(EISA), including the requirement that the combined mpg level of
U.S. cars and light trucks reach 35 miles per gallon by model year
2020. Because this provision would be expected to reduce future U.S.
demand for gasoline and lead to a decline in its future price, there
is some concern about whether the AEO 2010 forecast of fuel prices
partly reflects the increases in CAFE standards considered in this
rule, and thus whether it is suitable for valuing the projected
reductions in fuel use. In response to this concern, the agency
notes that EIA issued a revised version of AEO 2008 in June 2008,
which modified its previous December 2007 Early Release of AEO 2008
to reflect the effects of then recently-passed EISA legislation. The
fuel price forecasts reported in EIA's Revised Release of AEO 2008
differed by less than one cent per gallon throughout the entire
forecast period (2008-2030) from those previously issued as part of
its initial release of AEO 2008. Thus, the agencies are reasonably
confident that the fuel price forecasts presented in AEO 2010 and
used to analyze the value of fuel savings projected to result from
this rule are not unduly affected by the CAFE provisions of EISA.
---------------------------------------------------------------------------

While NHTSA relied on the forecasts of fuel prices presented in AEO
2008 High Price Case in the MY 2011 final rule, we noted at the time
that we were relying on that estimate primarily because volatility in
the oil market appeared to have overtaken the Reference Case. We also
anticipated that the Reference Case forecasts would be significantly
higher in subsequent editions of AEO, and that in future rulemaking
analyses the agency would be likely to rely on the Reference Case
rather than High Price Case forecasts. In fact, both EIA's AEO 2009
Reference Case and its subsequent AEO 2010 Early Release Reference Case
forecasts project higher retail fuel prices in most future years than
those forecast in the High Price Case from AEO 2008. NHTSA is thus
confident that the AEO 2010 Early Release Reference Case is an
appropriate forecast for projected future fuel prices.
NHTSA and EPA received relatively few comments on the fuel prices
used in the NPRM analysis, compared to previous CAFE rulemakings. Two
commenters, CARB and NADA, supported the use of AEO's Reference Case
for use in the agencies' analysis, although they disagreed on the
agencies' use of the High and Low Price Cases for sensitivities. Both
commenters emphasized the sensitivity of the market and the agencies'
analysis to higher and lower gas prices, and on that basis, CARB
supported the use of the High and Low Price Cases in sensitivity
analysis but urged the agencies to caveat the ``Reference Case''
results more explicitly. In contrast, NADA argued that the agencies
should not use the High and Low Price Cases, because EIA does not
assign specific probabilities to either of them. Only one commenter,
James Adcock, argued that the agencies should use forecasts of future
fuel prices other than those reported in AEO; Adcock stated that future
fuel prices should be assumed to be higher than current pump prices.
Measured in constant 2007 dollars, the AEO 2010 Early Release
Reference Case forecast of retail gasoline prices during calendar year
2010 is $2.44 per gallon, and rises gradually to $3.83 by the year 2035
(these values include Federal, State and local taxes). However, the
agency's analysis of the value of fuel savings over the lifetimes of MY
2012-2016 cars and light trucks requires forecasts extending through
calendar year 2050, approximately the last year during which a
significant number of MY 2016 vehicles will remain in service. To
obtain fuel price forecasts for the years 2036 through 2050, the agency
assumes that retail fuel prices will continue to increase after 2035 at
the average annual rates projected for 2025 through 2035 in the AEO
2010 Early Release Reference Case.\603\ This assumption results in a
projected retail price of gasoline that reaches $4.49 in 2007 dollars
during the year 2050.
---------------------------------------------------------------------------

\603\ This projection uses the rate of increase in fuel prices
for 2020-2030 rather than that over the complete forecast period
(2009-2030) because there is extreme volatility in the forecasts for
the years 2009 through approximately 2020. Using the average rate of
change over the complete 2009-2030 forecast period would result in
projections of declining fuel prices after 2030.
---------------------------------------------------------------------------

The value of fuel savings resulting from improved fuel economy to
buyers of light-duty vehicles is determined by the retail price of
fuel, which includes Federal, State, and any local taxes imposed on
fuel sales. The agency has updated the estimates of gasoline taxes it
employed in the NPRM using the recent data on State fuel tax rates;
expressed in 2007 dollars, Federal gasoline taxes are currently $0.178,
while State and local gasoline taxes together average $0.231 per
gallon, for a total tax burden of $0.401 per gallon. Because fuel taxes
represent transfers of resources from fuel buyers to government
agencies, however, rather than real resources that are consumed in the
process of supplying or using fuel, NHTSA deducts their value from
retail fuel prices to determine the true value of fuel savings
resulting from more stringent CAFE standards to the U.S. economy.
NHTSA follows the assumptions used by EIA in AEO 2010 Early Release
that State and local gasoline taxes will keep pace with inflation in
nominal terms, and thus remain constant when expressed in constant
dollars. In contrast, EIA assumes that Federal gasoline taxes will
remain unchanged in nominal terms, and thus decline throughout the
forecast period when expressed in constant dollars. These differing
assumptions about the likely future behavior of Federal and State/local
fuel taxes are consistent with recent historical experience, which
reflects the fact that Federal as well as most State motor fuel taxes
are specified on a cents-per-gallon rather than an ad valorem basis,
and typically require legislation to change. The projected value of
total taxes is deducted from each future year's forecast of retail
gasoline and diesel prices to determine the economic value of each
gallon of fuel saved during that year as a result of improved fuel
economy. Subtracting fuel taxes from the retail prices forecast in AEO
2010 Early Release results in a projected value for saving gasoline of
$2.04 per gallon during 2010, rising to $3.48 per gallon by the year
2035,and averaging $2.91 over this 25-year period.
Although the Early Release of AEO 2010 contains only the Reference
Case forecast, EIA includes ``High Price Case'' and ``Low Price Case''
forecasts in each year's complete AEO, which reflect uncertainties
regarding future levels of oil production and demand. For this final
rule, NHTSA has continued to use the most recent ``High Price Case''
and ``Low Price Case'' forecasts available, which are those from AEO
2009. While NHTSA recognizes that these forecasts are not
probabilistic, as NADA commented, we continue to believe that using
them for sensitivity analyses provides valuable information for agency
decision-makers, because it illustrates the sensitivity of the rule's
primary economic benefit resulting from uncertainty about future growth
in world demand for petroleum energy and the strategic behavior of oil
suppliers.
These alternative scenarios project retail gasoline prices that
range from a low of $2.02 to a high of $5.04 per gallon during 2020,
and from $2.04 to $5.47 per gallon during 2030 (all figures in 2007
dollars). In conjunction with our assumption that fuel taxes will
remain constant in real or inflation-adjusted terms over this period,
these forecasts imply pre-tax values of saving fuel ranging from $1.63
to $4.65 per gallon during 2020, and from $1.66 to $5.09 per gallon in
2030 (again, all figures are in constant 2007 dollars). In conducting
the analysis of uncertainty in benefits and costs from alternative CAFE
standards required by OMB, NHTSA evaluated the sensitivity of its
benefits estimates to these alternative forecasts of future fuel
prices. Detailed

[[Page 25585]]

results and discussion of this sensitivity analysis can be found in the
FRIA. Generally, however, this analysis confirmed that as several
commenters suggested, the primary economic benefit resulting from the
rule--the value of fuel savings--is quite sensitive to forecast fuel
prices.
e. Consumer Valuation of Fuel Economy and Payback Period
In estimating the impacts on vehicle sales that would result from
alternative CAFE standards to potential vehicle buyers, NHTSA assumes,
as in the MY 2011 final rule, that potential vehicle buyers value the
resulting fuel savings over only part of the expected lifetime of the
vehicles they purchase. Specifically, we assume that buyers value fuel
savings over the first five years of a new vehicle's lifetime, and
discount the value of these future fuel savings at a 3 percent annual
rate. The five-year figure represents approximately the current average
term of consumer loans to finance the purchase of new vehicles. We
recognize that the period over which individual buyers finance new
vehicle purchases may not correspond exactly to the time horizons they
apply in valuing fuel savings from higher fuel economy.
The agency deducts the discounted present value of fuel savings
over the first five years of a vehicle model's lifetime from the
technology costs incurred by its manufacturer to improve that model's
fuel economy to determine the increase in its ``effective price'' to
buyers. The Volpe model uses these estimates of effective costs for
increasing the fuel economy of each vehicle model to identify the order
in which manufacturers would be likely to select models for the
application of fuel economy-improving technologies in order to comply
with stricter standards. The average value of the resulting increase in
effective cost from each manufacturer's simulated compliance strategy
is also used to estimate the impact of alternative standards on its
total sales for future model years.
One commenter, NADA, supported the agency's assumption of a five-
year period for buyers' valuation of fuel economy, on the basis that
the considerable majority of consumers seek to recoup costs quickly.
However, NADA also encouraged the agencies to ensure that purchaser
finance costs, opportunity costs of vehicle ownership, and increased
maintenance costs were accounted for. Another commenter, James Adcock,
argued that the assumption of a five-year period was irrational,
because it did not account for the fact that first purchasers will be
able to sell a higher-mpg vehicle for more money than a lower-mpg
vehicle.
In response to these comments, the agency notes that it estimates
the aggregate value to the U.S. economy of fuel savings resulting from
alternative standards--or their ``social'' value--over the entire
expected lifetimes of vehicles manufactured under those standards,
rather than over the shorter 5-year ``payback period'' we assume that
manufacturers employ to represent the preferences of vehicle buyers.
The 5-year payback period is only utilized to identify the likely
sequence of improvements in fuel economy that manufacturers are likely
to make to their different vehicle models. The procedure the agency
uses for calculating lifetime fuel savings is discussed in detail in
the following section, while alternative assumptions about the time
horizon over which potential buyers consider fuel savings in their
vehicle purchasing decisions are analyzed and discussed in detail in
Section IV.G.6 below.
Valuing fuel savings over vehicles' entire lifetimes in effect
recognizes the gains that future vehicle owners will receive, even if
initial purchasers of higher-mpg models are not able to recover the
entire remaining value of fuel savings when they re-sell those
vehicles. The agency acknowledges, however, that it has not accounted
for any effects of increased financing costs for purchasing vehicles
with higher fuel economy or increased expenses for maintaining them on
benefits to vehicle owners, over either the short-run payback period or
the full lifetimes of vehicles.
f. Vehicle Survival and Use Assumptions
NHTSA's first step in estimating lifetime fuel consumption by
vehicles produced during a model year is to calculate the number
expected to remain in service during each year following their
production and sale.\604\ This is calculated by multiplying the number
of vehicles originally produced during a model year by the proportion
typically expected to remain in service at their age during each later
year, often referred to as a ``survival rate.''
---------------------------------------------------------------------------

\604\ Vehicles are defined to be of age 1 during the calendar
year corresponding to the model year in which they are produced;
thus for example, model year 2000 vehicles are considered to be of
age 1 during calendar year 2000, age 2 during calendar year 2001,
and to reach their maximum age of 26 years during calendar year
2025. NHTSA considers the maximum lifetime of vehicles to be the age
after which less than 2 percent of the vehicles originally produced
during a model year remain in service. Applying these conventions to
vehicle registration data indicates that passenger cars have a
maximum age of 26 years, while light trucks have a maximum lifetime
of 36 years. See Lu, S., NHTSA, Regulatory Analysis and Evaluation
Division, ``Vehicle Survivability and Travel Mileage Schedules,''
DOT HS 809 952, 8-11 (January 2006). Available at http://www-
nrd.nhtsa.dot.gov/Pubs/809952.pdf (last accessed March 1, 2010).
---------------------------------------------------------------------------

As discussed in more detail in Section II.B.3 above and in Chapter
1 of the TSD, to estimate production volumes of passenger cars and
light trucks for individual manufacturers, NHTSA relied on a baseline
market forecast constructed by EPA staff beginning with MY 2008 CAFE
certification data. After constructing a MY 2008 baseline, EPA and
NHTSA used projected car and truck volumes for this period from Energy
Information Administration's (EIA's) Annual Energy Outlook (AEO) 2009
in the NPRM analysis.\605\ For the analysis supporting this final rule,
NHTSA substituted the revised forecasts of total volume reported in
EIA's Annual Energy Outlook 2010 Early Release. However, Annual Energy
Outlook forecasts only total car and light truck sales, rather than
sales at the manufacturer and model-specific level, which the agencies
require in order to estimate the effects new standards will have on
individual manufacturers.\606\
---------------------------------------------------------------------------

\605\ Available at http://www.eia.doe.gov/oiaf/aeo/index.html
(last accessed March 15, 2010). NHTSA and EPA made the simplifying
assumption that projected sales of cars and light trucks during each
calendar year from 2012 through 2016 represented the likely
production volumes for the corresponding model year. The agency did
not attempt to establish the exact correspondence between projected
sales during individual calendar years and production volumes for
specific model years.
\606\ Because AEO 2009's ``car'' and ``truck'' classes did not
reflect NHTSA's recent reclassification (in March 2009 for
enforcement beginning MY 2011) of many two wheel drive SUVs from the
nonpassenger (i.e., light truck) fleet to the passenger car fleet,
EPA staff made adjustments to account for such vehicles in the
baseline.
---------------------------------------------------------------------------

To estimate sales of individual car and light truck models produced
by each manufacturer, EPA purchased data from CSM Worldwide and used
its projections of the number of vehicles of each type (car or truck)
that will be produced and sold by manufacturers in model years 2011
through 2015.\607\ This provided year-by-year estimates of the
percentage of cars and trucks sold by each manufacturer, as well as the
sales percentages accounted for by each vehicle market segment. (The
distributions of car and truck sales by manufacturer and by market
segment for the 2016 model year and beyond were assumed to be the same
as CSM's forecast for the 2015 calendar year.) Normalizing these
percentages to the

[[Page 25586]]

total car and light truck sales volumes projected for 2012 through 2016
in AEO 2009 provided manufacturer-specific market share and model-
specific sales estimates for those model years. The volumes were then
scaled to AEO 2010 total volume for each year.
---------------------------------------------------------------------------

\607\ EPA also considered other sources of similar information,
such as J.D. Powers, and concluded that CSM was better able to
provide forecasts at the requisite level of detail for most of the
model years of interest.
---------------------------------------------------------------------------

To estimate the number of passenger cars and light trucks
originally produced during model years 2012 through 2016 that will
remain in use during each subsequent year, the agency applied age-
specific survival rates for cars and light trucks to these adjusted
forecasts of passenger car and light truck sales. In 2008, NHTSA
updated its previous estimates of car and light truck survival rates
using the most current registration data for vehicles produced during
recent model years, in order to ensure that they reflected recent
increases in the durability and expected life spans of cars and light
trucks.\608\
---------------------------------------------------------------------------

\608\ Lu, S., NHTSA, Regulatory Analysis and Evaluation
Division, ``Vehicle Survivability and Travel Mileage Schedules,''
DOT HS 809 952, 8-11 (January 2006). Available at http://www-
nrd.nhtsa.dot.gov/Pubs/809952.pdf (last accessed March 1, 2010).
These updated survival rates suggest that the expected lifetimes of
recent-model passenger cars and light trucks are 13.8 and 14.5
years.
---------------------------------------------------------------------------

The next step in estimating fuel use is to calculate the total
number of miles that model year 2012-2016 cars and light trucks
remaining in use will be driven each year. To estimate total miles
driven, the number projected to remain in use during each future year
is multiplied by the average number of miles they are expected to be
driven at the age they will reach in that year. The agency estimated
annual usage of cars and light trucks of each age using data from the
Federal Highway Administration's 2001 National Household Transportation
Survey (NHTS).\609\ Because these estimates reflect the historically
low gasoline prices that prevailed at the time the 2001 NHTS was
conducted, however, NHTSA adjusted them to account for the effect on
vehicle use of subsequent increases in fuel prices. Details of this
adjustment are provided in Chapter VIII of the FRIA and Chapter 4 of
the Joint TSD.
---------------------------------------------------------------------------

\609\ For a description of the Survey, See http://nhts.ornl.gov/
quickStart.shtml (last accessed March 1, 2010).
---------------------------------------------------------------------------

Increases in average annual use of cars and light trucks have been
an important source of historical growth in the total number of miles
they are driven each year. To estimate future growth in their average
annual use for purposes of this rulemaking, NHTSA calculated the rate
of growth in the adjusted mileage schedules derived from the 2001 NHTS
necessary for total car and light truck travel to increase at the rate
forecast in the AEO 2010 Early Release Reference Case.\610\ This rate
was calculated to be consistent with future changes in the overall size
and age distributions of the U.S. passenger car and light truck fleets
that result from the agency's forecasts of total car and light truck
sales and updated survival rates. The resulting growth rate in average
annual car and light truck use of 1.15 percent per year was applied to
the mileage figures derived from the 2001 NHTS to estimate annual
mileage during each year of the expected lifetimes of MY 2012-2016 cars
and light trucks.\611\
---------------------------------------------------------------------------

\610\ This approach differs from that used in the MY 2011 final
rule, where it was assumed that future growth in the total number of
cars and light trucks in use resulting from projected sales of new
vehicles was adequate by itself to account for growth in total
vehicle use, without assuming continuing growth in average vehicle
use.
\611\ While the adjustment for future fuel prices reduces
average mileage at each age from the values derived from the 2001
NHTS, the adjustment for expected future growth in average vehicle
use increases it. The net effect of these two adjustments is to
increase expected lifetime mileage by about 18 percent significantly
for both passenger cars and about 16 percent for light trucks.
---------------------------------------------------------------------------

Finally, the agency estimated total fuel consumption by passenger
cars and light trucks remaining in use each year by dividing the total
number of miles surviving vehicles are driven by the fuel economy they
are expected to achieve under each alternative CAFE standard. Each
model year's total lifetime fuel consumption is the sum of fuel use by
the cars or light trucks produced during that model year during each
year of their life spans. In turn, the savings in a model year's
lifetime fuel use that will result from each alternative CAFE standard
is the difference between its lifetime fuel use at the fuel economy
level it attains under the Baseline alternative, and its lifetime fuel
use at the higher fuel economy level it is projected to achieve under
that alternative standard.\612\
---------------------------------------------------------------------------

\612\ To illustrate these calculations, the agency's adjustment
of the AEO 2009 Revised Reference Case forecast indicates that 9.26
million passenger cars will be produced during 2012, and the
agency's updated survival rates show that 83 percent of these
vehicles, or 7.64 million, are projected to remain in service during
the year 2022, when they will have reached an age of 10 years. At
that age, passenger achieving the fuel economy level they are
projected to achieve under the Baseline alternative are driven an
average of about 800 miles, so surviving model year 2012 passenger
cars will be driven a total of 82.5 billion miles (= 7.64 million
surviving vehicles x 10,800 miles per vehicle) during 2022. Summing
the results of similar calculations for each year of their 26-year
maximum lifetime, model year 2012 passenger cars will be driven a
total of 1,395 billion miles under the Baseline alternative. Under
that alternative, they are projected to achieve a test fuel economy
level of 32.4 mpg, which corresponds to actual on-road fuel economy
of 25.9 mpg (= 32.4 mpg x 80 percent). Thus their lifetime fuel use
under the Baseline alternative is projected to be 53.9 billion
gallons (= 1,395 billion miles divided by 25.9 miles per gallon).
---------------------------------------------------------------------------

NHTSA and EPA received no comments on their respective NPRMs
indicating that these assumptions should be updated or reconsidered.
Thus the agencies have continued to employ them in the analysis
supporting this final rule.
g. Accounting for the Fuel Economy Rebound Effect
The fuel economy rebound effect refers to the fraction of fuel
savings expected to result from an increase in vehicle fuel economy--
particularly an increase required by the adoption of higher CAFE
standards--that is offset by additional vehicle use. The increase in
vehicle use occurs because higher fuel economy reduces the fuel cost of
driving, typically the largest single component of the monetary cost of
operating a vehicle, and vehicle owners respond to this reduction in
operating costs by driving slightly more. By lowering the marginal cost
of vehicle use, improved fuel economy may lead to an increase in the
number of miles vehicles are driven each year and over their lifetimes.
Even with their higher fuel economy, this additional driving consumes
some fuel, so the rebound effect reduces the net fuel savings that
result when new CAFE standards require manufacturers to improve fuel
economy.
The magnitude of the rebound effect is an important determinant of
the actual fuel savings that are likely to result from adopting
stricter CAFE standards. Research on the magnitude of the rebound
effect in light-duty vehicle use dates to the early 1980s, and
generally concludes that a statistically significant rebound effect
occurs when vehicle fuel efficiency improves.\613\ The agency reviewed
studies of the rebound effect it had previously relied upon, considered
more recently published estimates, and developed new estimates of its
magnitude for purposes of the NPRM.\614\ Recent studies provide some
evidence that the rebound effect has been declining over time, and may
decline further over the immediate future if incomes rise faster than
gasoline prices. This result appears

[[Page 25587]]

plausible, because the responsiveness of vehicle use to variation in
fuel costs is expected to decline as they account for a smaller
proportion of the total monetary cost of driving, which has been the
case until very recently. At the same time, rising personal incomes
would be expected to reduce the sensitivity of vehicle use to fuel
costs as the time component of driving costs--which is likely to be
related to income levels--accounts for a larger fraction the total cost
of automobile travel.
---------------------------------------------------------------------------

\613\ Some studies estimate that the long-run rebound effect is
significantly larger than the immediate response to increased fuel
efficiency. Although their estimates of the adjustment period
required for the rebound effect to reach its long-run magnitude
vary, this long-run effect is most appropriate for evaluating the
fuel savings and emissions reductions resulting from stricter
standards that would apply to future model years.
\614\ For details of the agency's analysis, see Chapter VIII of
the PRIA and Chapter 4 of the draft Joint TSD accompanying this
proposed rule.
---------------------------------------------------------------------------

NHTSA developed new estimates of the rebound effect by using
national data on light-duty vehicle travel over the period from 1950
through 2006 to estimate various econometric models of the relationship
between vehicle miles-traveled and factors likely to influence it,
including household income, fuel prices, vehicle fuel efficiency, road
supply, the number of vehicles in use, vehicle prices, and other
factors.\615\ The results of NHTSA's analysis are consistent with the
findings from other recent research: the average long-run rebound
effect ranged from 16 percent to 30 percent over the period from 1950
through 2007, while estimates of the rebound effect in 2007 range from
8 percent to 14 percent. Projected values of the rebound effect for the
period from 2010 through 2030, which the agency developed using
forecasts of personal income, fuel prices, and fuel efficiency from AEO
2009's Reference Case, range from 4 percent to 16 percent, depending on
the specific model used to generate them.
---------------------------------------------------------------------------

\615\ The agency used several different model specifications and
estimation procedures to control for the effect of fuel prices on
fuel efficiency in order to obtain accurate estimates of the rebound
effect.
---------------------------------------------------------------------------

In light of these results, the agency's judgment is that the
apparent decline over time in the magnitude of the rebound effect
justifies using a value for future analysis that is lower than
historical estimates, which average 15-25 percent. Because the
lifetimes of vehicles affected by the alternative CAFE standards
considered in this rulemaking will extend from 2012 until nearly 2050,
a value that is significantly lower than historical estimates appears
to be appropriate. Thus NHTSA used a 10 percent rebound effect in its
analysis of fuel savings and other benefits from higher CAFE standards
for the NPRM. The agency also sought comment on other alternatives for
estimating the rebound effect, such as whether it would be appropriate
to use the price elasticity of demand for gasoline, or other
alternative approaches, to guide the choice of a value for the rebound
effect.
NHTSA and EPA received far fewer comments on the rebound effect
than were previously received to CAFE rulemakings. Only one commenter,
NJ DEP, expressly supported the agencies' assumption of 10 percent for
the rebound effect; other commenters (CARB, CBD, ICCT) argued that 10
percent should be the absolute maximum value and that the rebound
effect assumed by the agencies should be lower, and would also be
expected to decline over time. ICCT added that the price elasticity of
gasoline demand could be a useful comparison for the rebound effect,
but should not be used to derive it. Other commenters argued that a
rebound effect either was unlikely to occur (James Hyde), or was
unlikely to produce a uniform increase in use of all vehicles with
improved fuel economy (Missouri DNR). NADA argued, in contrast, that
the agencies had not provided sufficient justification for lowering the
rebound effect to 10 percent from the ``historically justified'' range
of 15 to 30 percent.
The agency's interpretation of historical and recent evidence on
the magnitude of the rebound effect is that a significant fuel economy
rebound effect exists, and commenters did not provide any additional
data or analysis to justify revising our initial estimates of the
rebound effect. Therefore, the data available at this time do not
justify using a rebound effect below the 10 percent figure employed in
its NPRM analysis. NHTSA believes that projections of a continued
decline in the magnitude of the rebound effect are unrealistic because
they assume the rate at which it declines in response to increasing
incomes remain constant, and in some cases imply that the rebound
effect will become negative in the near future. In addition, the
continued increases in fuel prices used in this analysis will tend to
increase the magnitude of the rebound effect, thus offsetting part of
the effect of rising incomes. As the preceding discussion indicates,
there is a wide range of estimates for both the historical magnitude of
the rebound effect and its projected future value, and there is some
evidence that the magnitude of the rebound effect appears to be
declining over time. Nevertheless, NHTSA requires a single point
estimate for the rebound effect as an input to its analysis, although a
range of estimates can be used to test the sensitivity to uncertainty
about its exact magnitude. For the final rule, NHTSA chose to use 10
percent as its primary estimate of the rebound effect, with a range of
5-15 percent for use in sensitivity testing.
The 10 percent figure is well below those reported in almost all
previous research, and it is also below most estimates of the
historical and current magnitude of the rebound effect developed by
NHTSA. However, other recent research--particularly that conducted by
Small and Van Dender and by Greene--reports persuasive evidence that
the magnitude of the rebound effect is likely to be declining over
time, and the forecasts developed by NHTSA also suggest that this is
likely to be the case. As a consequence, NHTSA concluded that a value
below the historical estimates reported here is likely to provide a
more reliable estimate of its magnitude during the future period
spanned by NHTSA's analysis of the impacts of this rule. The 10 percent
estimate meets this condition, since it lies below the 15-30 percent
range of estimates for the historical rebound effect reported in most
previous research, and at the upper end of the 5-10 percent range of
estimates for the future rebound effect reported in the recent studies
by Small and Van Dender and by Greene. It also lies within the 3-16
percent range of forecasts of the future magnitude of the rebound
effect developed by NHTSA in its recent research. In summary, the 10
percent value was not derived from a single point estimate from a
particular study, but instead represents a reasonable compromise
between the historical estimates and the projected future estimates.
NHTSA will continue to review this estimate of the rebound effect in
future rulemakings, but the agency has continued to use the 10 percent
rebound effect over the entire future period spanned by the analysis it
conducted for this final rule.
h. Benefits From Increased Vehicle Use
The increase in vehicle use from the rebound effect provides
additional benefits to their owners, who may make more frequent trips
or travel farther to reach more desirable destinations. This additional
travel provides benefits to drivers and their passengers by improving
their access to social and economic opportunities away from home. As
evidenced by their decisions to make more frequent or longer trips when
improved fuel economy reduces their costs for driving, the benefits
from this additional travel exceed the costs drivers and passengers
incur in making more frequent or longer trips.
The agency's analysis estimates the economic benefits from
increased rebound-effect driving as the sum of fuel costs drivers incur
plus the consumer surplus they receive from the additional

[[Page 25588]]

accessibility it provides.\616\ Because the increase in travel depends
on the extent of improvement in fuel economy, the value of benefits it
provides differs among model years and alternative CAFE standards.
Under even those alternatives that would impose the highest standards,
however, the magnitude of these benefits represents a small fraction of
total benefits. Because no comments addressed this issue of benefits
from increased vehicle use or the procedure used to estimate them, the
agencies have finalized their proposed assumptions for purposes of the
final rule analysis.
---------------------------------------------------------------------------

\616\ The consumer surplus provided by added travel is estimated
as one-half of the product of the decline in fuel cost per mile and
the resulting increase in the annual number of miles driven.
---------------------------------------------------------------------------

i. The Value of Increased Driving Range
Improving vehicles' fuel economy may also increase their driving
range before they require refueling. By reducing the frequency with
which drivers typically refuel, and by extending the upper limit of the
range they can travel before requiring refueling, improving fuel
economy thus provides some additional benefits to their owners.\617\
NHTSA re-examined this issue for purposes of this rulemaking, and found
no information in comments or elsewhere that would cause the agency to
revise its previous approach. Since no direct estimates of the value of
extended vehicle range are available, NHTSA calculates directly the
reduction in the annual number of required refueling cycles that
results from improved fuel economy, and applies DOT-recommended values
of travel time savings to convert the resulting time savings to their
economic value.\618\
---------------------------------------------------------------------------

\617\ If manufacturers respond to improved fuel economy by
reducing the size of fuel tanks to maintain a constant driving
range, the resulting cost saving will presumably be reflected in
lower vehicle sales prices.
\618\ See Department of Transportation, Guidance Memorandum,
``The Value of Saving Travel Time: Departmental Guidance for
Conducting Economic Evaluations,'' Apr. 9, 1997. http://
ostpxweb.dot.gov/policy/Data/VOT97guid.pdf (last accessed March 1,
2010); update available at http://ostpxweb.dot.gov/policy/Data/
VOTrevision1_2-11-03.pdf (last accessed March 1, 2010).
---------------------------------------------------------------------------

As an illustration, a typical small light truck model has an
average fuel tank size of approximately 20 gallons. Assuming that
drivers typically refuel when their tanks are 55 percent full (i.e., 11
gallons in reserve), increasing this model's actual on-road fuel
economy from 24 to 25 mpg would extend its driving range from 216 miles
(= 9 gallons x 24 mpg) to 225 miles (= 9 gallons x 25 mpg). Assuming
that it is driven 12,000 miles/year, this reduces the number of times
it needs to be refueled each year from 55.6 (= 12,000 miles per year/
216 miles per refueling) to 53.3 (= 12,000 miles per year/225 miles per
refueling), or by 2.3 refuelings per year.
Weighted by the nationwide mix of urban and rural driving, personal
and business travel in urban and rural areas, and average vehicle
occupancy for driving trips, the DOT-recommended values of travel time
per vehicle-hour is $24.64 (in 2007 dollars).\619\Assuming that
locating a station and filling up requires a total of five minutes, the
annual value of time saved as a result of less frequent refueling
amounts to $4.72 (calculated as 5/60 x 2.3 x $24.64). This calculation
is repeated for each future year that model year 2012-2016 cars and
light trucks would remain in service. Like fuel savings and other
benefits, the value of this benefit declines over a model year's
lifetime, because a smaller number of vehicles originally produced
during that model year remain in service each year, and those remaining
in service are driven fewer miles.
---------------------------------------------------------------------------

\619\ The hourly wage rate during 2008 is estimated to average
$25.50 when expressed in 2007 dollars. Personal travel in urban
areas (which represents 94 percent of urban travel) is valued at 50
percent of the hourly wage rate, while business travel (the
remaining 6 percent of urban travel) is valued at 100 percent of the
hourly wage rate. For intercity travel, personal travel (87 percent
of total intercity travel) is valued at 70 percent of the wage rate,
while business travel (13 percent) is valued at 100 percent of the
wage rate. The resulting values of travel time are $12.67 for urban
travel and $17.66 for intercity travel, and must be multiplied by
vehicle occupancy (1.6) to obtain the estimated values of time per
vehicle hour in urban and rural driving. Finally, about 66% of
driving occurs in urban areas, while the remaining 34% takes place
in rural areas, and these percentages are used to calculate a
weighted average of the value of time in all driving.
---------------------------------------------------------------------------

Although the agencies received no public comments on the procedures
they used to estimate the benefits from less frequent refueling or the
magnitude of those benefits, we note also that the estimated value of
less frequent refueling events is subject to a number of uncertainties
which we discuss in detail in Chapter 4.1.11 of the Joint TSD, and the
actual value could be higher or lower than the value presented here.
Specifically, the analysis makes three assumptions: (a) That
manufacturers will not adjust fuel tank capacities downward (from the
current average of 19.3 gallons) when they improve the fuel economy of
their vehicle models. (b) that the average fuel purchase (55 percent of
fuel tank capacity) is the typical fuel purchase. (c) that 100 percent
of all refueling is demand-based; i.e., that every gallon of fuel which
is saved would reduce the need to return to the refueling station.
NHTSA has planned a new research project which will include a detailed
study of refueling events, and which is expected to improve upon these
assumptions. These assumptions and the upcoming research project are
discussed in detail in Joint TSD Chapter 4.2.10, as well as in Chapter
VIII of NHTSA's FRIA.
j. Added Costs From Congestion, Crashes and Noise
Increased vehicle use associated with the rebound effect also
contributes to increased traffic congestion, motor vehicle accidents,
and highway noise. NHTSA relies on estimates of per-mile congestion,
accident, and noise costs caused by increased use of automobiles and
light trucks developed by the Federal Highway Administration to
estimate these increased costs.\620\ NHTSA employed these estimates
previously in its analysis accompanying the MY 2011 final rule, and
after reviewing the procedures used by FHWA to develop them and
considering other available estimates of these values, continues to
find them appropriate for use in this final rule. The agency multiplies
FHWA's estimates of per-mile costs by the annual increases in
automobile and light truck use from the rebound effect to yield the
estimated increases in congestion, accident, and noise externality
costs during each future year.
---------------------------------------------------------------------------

\620\ These estimates were developed by FHWA for use in its 1997
Federal Highway Cost Allocation Study; See http://www.fhwa.dot.gov/
policy/hcas/final/index.htm (last accessed March 1, 2010).
---------------------------------------------------------------------------

One commenter, Inrix, Inc., stated that ``deeply connected
vehicles,'' i.e., those with built-in computer systems to help drivers
identify alternative routes to avoid congestion, are better able to
avoid congestion than conventional vehicles. The commenter argued that
increased use of these models may be less likely to contribute to
increased congestion, and urged the agencies to consider the impact of
this on their estimates of fuel use and GHG emissions. NHTSA notes that
the number of such vehicles is extremely small at present, and is
likely to remain modest for the model years affected by this rule, and
has thus continued to employ the estimates of congestion costs from
additional rebound-effect vehicle use that it utilized in the NPRM
analysis. The agency recognizes that these vehicles may become
sufficiently common in the future that their effect on the fuel economy
drivers actually experience could become significant, but notes that to
the extent this occurs,

[[Page 25589]]

it would be reflected in the gap between test and on-road fuel economy.
NHTSA will continue to monitor the production of such vehicles and
their representation in the vehicle fleet in its future rulemakings.
k. Petroleum Consumption and Import Externalities
U.S. consumption and imports of petroleum products also impose
costs on the domestic economy that are not reflected in the market
price for crude petroleum, or in the prices paid by consumers of
petroleum products such as gasoline. These costs include (1) higher
prices for petroleum products resulting from the effect of U.S. oil
import demand on the world oil price; (2) the risk of disruptions to
the U.S. economy caused by sudden reductions in the supply of imported
oil to the U.S.; and (3) expenses for maintaining a U.S. military
presence to secure imported oil supplies from unstable regions, and for
maintaining the strategic petroleum reserve (SPR) to cushion against
resulting price increases.\621\
---------------------------------------------------------------------------

\621\ See, e.g., Bohi, Douglas R. and W. David Montgomery
(1982). Oil Prices, Energy Security, and Import Policy Washington,
DC: Resources for the Future, Johns Hopkins University Press; Bohi,
D.R., and M.A. Toman (1993). ``Energy and Security: Externalities
and Policies,'' Energy Policy 21:1093-1109 (Docket NHTSA-2009-0062-
24); and Toman, M.A. (1993). ``The Economics of Energy Security:
Theory, Evidence, Policy,'' in A.V. Kneese and J.L. Sweeney, eds.
(1993) (Docket NHTSA-2009-0062-23). Handbook of Natural Resource and
Energy Economics, Vol. III. Amsterdam: North-Holland, pp. 1167-1218.
---------------------------------------------------------------------------

Higher U.S. imports of crude oil or refined petroleum products
increase the magnitude of these external economic costs, thus
increasing the true economic cost of supplying transportation fuels
above their market prices. Conversely, lowering U.S. imports of crude
petroleum or refined fuels by reducing domestic fuel consumption can
reduce these external costs, and any reduction in their total value
that results from improved fuel economy represents an economic benefit
of more stringent CAFE standards, in addition to the value of saving
fuel itself.
NHTSA has carefully reviewed its assumptions regarding the
appropriate value of these benefits for this final rule. In analyzing
benefits from its recent actions to increase light truck CAFE standards
for model years 2005-07 and 2008-11, NHTSA relied on a 1997 study by
Oak Ridge National Laboratory (ORNL) to estimate the value of reduced
economic externalities from petroleum consumption and imports.\622\
More recently, ORNL updated its estimates of the value of these
externalities, using the analytic framework developed in its original
1997 study in conjunction with recent estimates of the variables and
parameters that determine their value.\623\ The updated ORNL study was
subjected to a detailed peer review comissioned by EPA, and ORNL's
estimates of the value of oil import externalities were subsequently
revised to reflect their comments and recommendations of the peer
reviewers.\624\ Finally, at the request of EPA, ORNL further revised
its 2008 estimates of external costs from U.S. oil imports to reflect
recent changes in the outlook for world petroleum prices, as well as
continuing changes in the structure and characteristics of global
petroleum supply and demand.
---------------------------------------------------------------------------

\622\ Leiby, Paul N., Donald W. Jones, T. Randall Curlee, and
Russell Lee, Oil Imports: An Assessment of Benefits and Costs, ORNL-
6851, Oak Ridge National Laboratory, November 1, 1997. Available at
http://pzl1.ed.ornl.gov/ORNL6851.pdf (last accessed March 1, 2010).
\623\ Leiby, Paul N. ``Estimating the Energy Security Benefits
of Reduced U.S. Oil Imports,'' Oak Ridge National Laboratory, ORNL/
TM-2007/028, Revised July 23, 2007. Available at http://
pzl1.ed.ornl.gov/energysecurity.html (click on link below ``Oil
Imports Costs and Benefits'') (last accessed March 1, 2010).
\624\ Peer Review Report Summary: Estimating the Energy Security
Benefits of Reduced U.S. Oil Imports, ICF, Inc., September 2007.
Available at Docket No. NHTSA-2009-0059-0160.
---------------------------------------------------------------------------

These most recent revisions increase ORNL's estimates of the
``monopsony premium'' associated with U.S. oil imports, which measures
the increase in payments from U.S. oil purchasers to foreign oil
suppliers beyond the increased purchase price of petroleum itself that
results when increased U.S. import demand raises the world price of
petroleum.\625\ However, the monopsony premium represents a financial
transfer from consumers of petroleum products to oil producers, which
does not entail the consumption of real economic resources. Thus
reducing the magnitude of the monopsony premium produces no savings in
real economic resources globally or domestically, although it does
reduce the value of the financial transfer from U.S. consumers of
petroleum products to foreign suppliers of petroleum. Accordingly,
NHTSA's analysis of the benefits from adopting proposed CAFE standards
for MY 2012-2016 cars and light trucks excluded the reduced value of
monopsony payments by U.S. oil consumers that might result from lower
fuel consumption by these vehicles. The agency sought comment on
whether it would be reasonable to include the reduction in monopsony
payments by U.S. consumers of petroleum products in their estimates of
total economic benefits from reducing U.S. fuel consumption.
---------------------------------------------------------------------------

\625\ The reduction in payments from U.S. oil purchasers to
domestic petroleum producers is not included as a benefit, since it
represents a transfer that occurs entirely within the U.S. economy.
---------------------------------------------------------------------------

Commenters from NYU School of Law argued that monopsony payments
should be treated as a distributional effect, not a standard efficiency
benefit. An individual commenter, A.G. Fraas, also supported the
agencies' exclusion of the monopsony benefit, arguing that it
represents a pecuniary externality that should not be considered in
benefit-cost analyses of governmental actions--again, in essence, that
it represents a distributional effect. These comments support the
agency's decision to exclude any reduction in monopsony premium
payments that results from lower U.S. petroleum imports from its
accounting of benefits from reduced fuel consumption. Thus the agency
continues to exclude any reduction in monopsony premium payments from
its estimates of benefits for the stricter CAFE standards this final
rule establishes.
ORNL's most recently revised estimates of the increase in the
expected costs associated with potential disruptions in U.S. petroleum
imports imply that each gallon of imported fuel or petroleum saved
reduces the expected costs of oil supply disruptions to the U.S.
economy by $0.169 per gallon (in 2007$). In contrast to reduced
monopsony premium payments, the reduction in expected disruption costs
represents a real savings in resources, and thus contributes economic
benefits in addition to the savings in fuel production costs that
result from increasing fuel economy. NHTSA employs this value in its
analysis of the economic benefits from adopting higher CAFE standards
for MY 2012-2016 cars and light trucks.
A.G. Fraas commented on this proposed rule and felt that that
magnitude of the economic disruption portion of the energy security
benefit may be too high. He cites a recent paper written by Stephen
P.A. Brown and Hillard G. Huntington, entitled ``Estimating U.S. Oil
Security Premiums'' (September 2009). He commented that the Brown and
Huntington premium associated with replacing oil imports by increased
domestic oil production while keeping U.S. oil consumption unchanged
(i.e., ``the cost of displacing a barrel of domestic oil with a barrel
of imported oil'') ranges from $2.17 per barrel in 2015 to $2.37 per
barrel in 2030 (2007$), or $0.052 to $0.056 per gallon.
In contrast, this rule is not a domestic oil supply initiative, but
is one intended to reduce domestic oil consumption and thereby also to
a significant extent reduce U.S. oil imports. When NHTSA

[[Page 25590]]

used the ORNL Energy Security Premium Analysis to calculate the energy
security premium for this rule, it based the energy security premium on
decreased demand for oil and oil products. The agency estimated that
most of the decreased demand for oil and oil products would come from
decreased imports of oil, given the inelasticity of U.S. supply and the
modest estimated change in world oil price. The Brown and Huntington
estimates for this change, considering the disruption component alone,
are much in line with the ORNL estimates. For a reduction in U.S.
consumption that largely leads to a reduction in imports, Brown and
Huntington estimate a midpoint premium of $4.98 per barrel in 2015
rising to $6.82 per barrel by 2030 (2007$). The 2015 disruption premium
estimate has an uncertainty range of $1.10 to $14.35 (2007$). The
corresponding 2030 estimate from ORNL is only about 19 percent higher
($8.12/bbl), with an uncertainty range--$3.90 to $13.04--completely
enclosed by that of Brown and Huntington. Thus, we conclude that the
ORNL disruption security premium estimates for this rule is roughly
consistent with the Brown and Huntington results.
Commenters from the NYU School of Law agreed that reduced
disruption costs should be counted as a benefit, but stated that the
agencies should disaggregate and exclude any reduction in wealth
transfers that occur during oil shocks from their calculation of this
benefit. NHTSA acknowledges that for consistency with its exclusion of
reductions in monopsony premium payments from the benefits of reduced
fuel consumption and petroleum imports, it may be necessary to exclude
reductions in the wealth transfer component of macroeconomic disruption
costs from the benefits of reducing U.S. petroleum imports. In future
rulemakings, the agency will assess the arguments for excluding the
wealth transfer component of disruption costs from its accounting of
benefits from reducing domestic fuel consumption and U.S. petroleum
imports, and explore whether it is practical to estimate its value
separately and exclude it from the benefits calculations.
NHTSA's analysis does not include savings in budgetary outlays to
support U.S. military activities among the benefits of higher fuel
economy and the resulting fuel savings.\626\ NHTSA's analysis of
benefits from alternative CAFE standards for MY 2012-2016 also excludes
any cost savings from maintaining a smaller SPR from its estimates of
the external benefits of reducing gasoline consumption and petroleum
imports. This view concurs with that of the recent ORNL study of
economic costs from U.S. oil imports, which concludes that savings in
government outlays for these purposes are unlikely to result from
reductions in consumption of petroleum products and oil imports on the
scale of those resulting from higher CAFE standards.
---------------------------------------------------------------------------

\626\ However, the agency conducted a sensitivity analysis of
the potential effect of assuming that some reduction military
spending would result from fuel savings and reduced petroleum
imports in order to investigate its impacts on the standards and
fuel savings.
---------------------------------------------------------------------------

Commenters from the NYU School of Law stated that the agencies were
justified in not including a value for military security, as long as
the agencies incorporate the increased protection value of the SPR into
their calculation of disruption effects. CBD and James Adcock
disagreed, and stated that the agencies should, in fact, include a
value for military security--CBD cited several studies, and Mr. Adcock
presented his own value of $0.275 per gallon. CARB stated simply that
the agencies should include a sensitivity analysis for military
security at $0.15 per gallon, in addition to the $0.05 per gallon
already evaluated. EDF also cited studies claiming a benefit for
increased national security.
In response to the comments from CBD and Mr. Adcock, NHTSA's
examination of the historical record indicates that while costs for
U.S. military security may vary over time in response to long-term
changes in the level of oil imports into the U.S., these costs are
unlikely to decline in response to the small reductions in U.S. oil
imports (relative to total oil imports) that are typically projected to
result from raising CAFE standards for light-duty vehicles. U.S.
military activities in regions that represent vital sources of oil
imports also serve a broader range of security and foreign policy
objectives than simply protecting oil supplies, and as a consequence
are unlikely to vary significantly in response to the modest changes in
the level of oil imports likely to be prompted by higher CAFE
standards.
The agency does not find evidence in the historical record that
Congress or the Executive Branch has ever attempted to calibrate U.S.
military expenditures, overall force levels, or specific deployments to
any measure of global oil market activity or U.S. reliance on petroleum
imports, or to any calculation of the projected economic consequences
of hostilities arising in the Persian Gulf. Instead, changes in U.S.
force levels, deployments, and thus military spending in that region
have been largely governed by political events, emerging threats, and
other military and political considerations, rather than by shifts in
U.S. oil consumption or imports. NHTSA thus concludes that the levels
of U.S. military activity and expenditures are likely to remain
unaffected by even relatively large changes in light duty vehicle fuel
consumption, and has continued to exclude any reduction in these
outlays from its estimates of the economic benefits resulting from
lower U.S. fuel consumption and petroleum imports.
In response to the comments from the NYU School of Law, NHTSA will
explore how it might estimate the contribution of the SPR to reducing
potential macroeconomic costs from oil supply disruptions, although the
agency notes that to some extent the existence of the SPR may already
be reflected in the magnitude of price elasticities of the supplies of
foreign oil available for import to the U.S. However, the agency notes
that the size of the SPR has not appeared to change significantly in
response to historical variation in U.S. petroleum consumption or
imports, suggesting that its effect on the magnitude of potential
macroeconomic costs from disruptions in petroleum imports may be
limited.
Finally, in response to the comment from EDF, the agency notes that
the value of $0.05 per gallon for the reduction in military security
outlays that is used for sensitivity analysis assumes that the entire
reduction in U.S. petroleum imports resulting from higher CAFE
standards would reflect lower imports from Persian Gulf suppliers, that
the estimate of annual U.S. military costs for securing Persian Gulf
oil supplies reported by Delucchi and Murphy is correct, and that
Congress would reduce half of these outlays in proportion to any
decline in U.S. oil imports from the region. The $0.15 per gallon
estimate recommended by CARB would thus require that U.S. military
outlays to protect Persian Gulf oil supplies are three times as large
as Delucchi and Murphy estimate, or that Congress would reduce military
spending in that region more than in proportion to any reduction in
U.S. petroleum imports originating there. Because it views these
possibilities as unrealistic, NHTSA has continued to use the $0.05
figure in its sensitivity analysis, rather than the higher figure
suggested.
Based on a detailed analysis of differences in fuel consumption,

[[Page 25591]]

petroleum imports, and imports of refined petroleum products among the
Reference Case, High Economic Growth, and Low Economic Growth Scenarios
presented in AEO 2009, NHTSA estimated that approximately 50 percent of
the reduction in fuel consumption resulting from adopting higher CAFE
standards is likely to be reflected in reduced U.S. imports of refined
fuel, while the remaining 50 percent would reduce domestic fuel
refining.\627\ Of this latter figure, 90 percent is anticipated to
reduce U.S. imports of crude petroleum for use as a refinery feedstock,
while the remaining 10 percent is expected to reduce U.S. domestic
production of crude petroleum.\628\ Thus on balance, each 100 gallons
of fuel saved as a consequence of higher CAFE standards is anticipated
to reduce total U.S. imports of crude petroleum or refined fuel by 95
gallons.\629\
---------------------------------------------------------------------------

\627\ Differences between forecast annual U.S. imports of crude
petroleum and refined products among these three scenarios range
from 24-89 percent of differences in projected annual gasoline and
diesel fuel consumption in the U.S. These differences average 49
percent over the forecast period spanned by AEO 2009.
\628\ Differences between forecast annual U.S. imports of crude
petroleum among these three scenarios range from 67-97 percent of
differences in total U.S. refining of crude petroleum, and average
85 percent over the forecast period spanned by AEO 2009.
\629\ This figure is calculated as 50 gallons + 50 gallons*90% =
50 gallons + 45 gallons = 95 gallons.
---------------------------------------------------------------------------

NHTSA employed this estimate in the analysis presented in the NPRM,
and received no comments on the assumptions or data used to develop it.
Hence the agency has continued to assume that each 100 gallons of fuel
saved as a consequence of the CAFE standards established by this final
rule will reduce total U.S. imports of crude petroleum or refined fuel
by 95 gallons. NHTSA has applied the estimates of economic benefits
from lower U.S. petroleum imports to the resulting estimate of
reductions in imports of crude petroleum and refined fuel.
l. Air Pollutant Emissions
i. Changes in Criteria Air Pollutant Emissions
Criteria air pollutants emitted by vehicles and during fuel
production include carbon monoxide (CO), hydrocarbon compounds (usually
referred to as ``