[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 are not available.
    EPA believes that the useful life period established for criteria 
pollutants under Tier 2 is also appropriate for CO2. Data 
from EPA's current in-use compliance test program indicate that 
CO2 emissions from current technology vehicles increase very 
little with age and in some cases may actually improve slightly. The 
stable CO2 levels are expected because unlike criteria 
pollutants, CO2 emissions in current technology vehicles are 
not controlled by after treatment systems that may fail with age. 
Rather, vehicle CO2 emission levels depend primarily on 
fundamental vehicle design characteristics that do not change over 
time. Therefore, vehicles designed for a given CO2 emissions 
level will be expected to sustain the same emissions profile over their 
full useful life.
    The CAA requires emission standards to be applicable for the 
vehicle's full useful life. Under Tier 2 and other vehicle emission 
standard programs, EPA requires manufacturers to demonstrate at the 
time of certification that the new vehicles being certified will 
continue to meet emission standards throughout their useful life. EPA 
allows manufacturers several options for predicting in-use 
deterioration, including full vehicle testing, bench-aging specific 
components, and application of a deterioration factor based on data 
and/or engineering judgment.
    In the specific case of CO2, EPA does not currently 
anticipate notable deterioration and has therefore determined that an 
assigned deterioration factor be applied at the time of certification. 
At this time EPA will use an additive assigned deterioration factor of 
zero, or a multiplicative factor of one. EPA anticipates that the 
deterioration factor will be updated from time to time, as new data 
regarding emissions deterioration for CO2 are obtained and 
analyzed. Additionally, EPA may consider technology-specific 
deterioration factors, should data indicate that certain CO2 
control technologies deteriorate differently than others.
    During compliance plan discussions prior to the beginning of the 
certification process, EPA will explore with each manufacturer any new 
technologies that could warrant use of a different deterioration 
factor. For any vehicle model determined likely to experience increases 
in CO2 emissions over the vehicle's useful life, 
manufacturers will not be allowed to use the assigned deterioration 
factor but rather will be required to establish an appropriate factor. 
If such an instance were to occur, EPA would allow manufacturers to use 
the whole-vehicle mileage accumulation method currently offered in 
EPA's regulations.\277\
---------------------------------------------------------------------------

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

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

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

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

[[Page 25475]]

in-use CO2 results, which are based on the combination of 
FTP and highway cycle test results, the highway fuel economy test would 
also need to be part of IUVP. The University of California, Santa 
Barbara expressed support for including N2O and 
CH4 emissions as part of the IUVP. Manufacturer comments 
were almost unanimously opposed to including any GHG as part of the 
IUVP. Specifically, industry commented that CO2 emissions do 
not deteriorate over time and in some cases actually improve. Ford 
provided data for several 2004 through 2007 model year vehicles that 
indicate CO2 emissions improved an average of 1.42% when 
vehicles wer