Federal Register, Volume 75 Issue 88 (Friday, May 7, 2010)

[Federal Register Volume 75, Number 88 (Friday, May 7, 2010)]
[Rules and Regulations]
[Pages 25323-25728]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 2010-8159]

[[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

Federal Register / Vol. 75, No. 88 / Friday, May 7, 2010 / Rules and
Regulations

[[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