Draft Regulatory Impact Analysis

   Proposed Rulemaking to Establish Light-
   Duty Vehicle Greenhouse Gas Emission
   Standards and Corporate Average Fuel
   Economy Standards
United States
Environmental Protection
Agency

-------
                Draft Regulatory Impact Analysis

             Proposed Rulemaking to Establish Light-
             Duty Vehicle Greenhouse Gas Emission
              Standards and Corporate Average Fuel
                        Economy Standards
                         Assessment and Standards Division
                        Office of Transportation and Air Quality
                        U.S. Environmental Protection Agency
v>EPA
United States                               EPA-420-D-09-003
Environmental Protection                          _  ^ ,  „„_
Agency                                  September 2009

-------
                                                Draft Regulatory Impact Analysis
                                  Table of Contents
TABLE OF CONTENTS	i
LIST OF ACRONYMS	vi
EXECUTIVE SUMMARY	ix
Greenhouse Emission Impacts of EPA's Proposal	x
Criteria Pollutant Impacts of EPA's Proposal	xi
Costs and Benefits of EPA's Proposal	xiii
CHAPTER 1:   TECHNOLOGY PACKAGES, COST AND EFFECTIVENESSl-1
1.1   Overview of Technology	1-1
1.2   Technology Cost and Effectiveness	1-4
1.3   Package Cost and Effectiveness	1-10
  1.3.1    Explanation of Technology Packages	1-10
  1.3.2    Technology Package Costs & Effectiveness	1-12
1.4   EPA's Lumped Parameter Approach for Determining Effectiveness Synergies. 1-26
  1.4.1    Ricardo's Vehicle Simulation	1-28
  1.4.2    Description of Ricardo's Report	1-29
  1.4.3    Determination of representative vehicle classes	1-30
  1.4.4    Description of Baseline Vehicle Models	1-31
  1.4.5    Technologies Considered by EPA and Ricardo in the Vehicle Simulation 1-32
  1.4.6    Choice of Technology Packages	1-34
  1.4.7    Simulation Results	1-35
1.5   Comparison of Lumped-Parameter Results to Modeling Results	1-36
1.6   Using the Lumped-Parameter Technique to Determine Synergies in a Technology
Application Flowpath (Identifying "Technology Pairs" to account for synergies)	1-38
CHAPTER 2:   AIR CONDITIONING	2-2
2.1   Overview of Air Conditioning Impacts and Technologies	2-2
2.2   Air Conditioner Leakage	2-4
  2.2.1    Impacts of Refrigerant Leakage on Greenhouse Gas Emissions	2-4
  2.2.2    A/C Leakage Credit	2-5
  2.2.3    Technologies That Reduce Refrigerant Leakage and their Effectiveness..2-ll
  2.2.4    Technical Feasibility of Leakage-Reducing Technologies	2-14
  2.2.5    Deterioration of Leakage Controls in A/C Systems	2-14
  2.2.6    Other Benefits of improving A/C Leakage Performance	2-18

-------
Draft Regulatory Impact Analysis
2.3   COi Emissions due to Air Conditioners	2-19
  2.3.1    Impact of Air Conditioning Use on Fuel Consumption and CO2 Emissions.2-
  19
  2.3.2    Technologies That Improve Efficiency of Air Conditioning and Their
  Effectiveness	2-28
  2.3.3    Technical Feasibility of Efficiency-Improving Technologies	2-32
  2.3.4    A/C Efficiency Credits	2-32
2.4   Costs of A/C reducing technologies	2-37
2.5   Air Conditioning Credit Summary	2-39
CHAPTER 3:  TECHNICAL BASIS OF THE STANDARDS	3-1
3.1   Technical Basis of the Standards	3-1
  3.1.1    Summary	3-1
  3.1.2    Overview of Equivalency Calculation	3-2
3.2   Analysis of Footprint Approach for Establishing Individual Company Standards 3-
7
  3.2.1    "Footprint" as a vehicle attribute	3-8
  3.2.2    Alternative Attributes	3-14
  3.2.3    EPA Selection of the Footprint Attribute	3-20
3.3   Supplemental Analysis of Relative Car and Truck Standards	3-23
CHAPTER 4:  RESULTS OF PROPOSED AND ALTERNATIVE STANDARDS  4-1
4.1   Introduction	4-1
4.2   Model Inputs	4-1
  4.2.1    Representation of the CO2 Control Technology Already Applied to 2008 MY
  Vehicles 4-2
  4.2.2    Technology Package Approach	4-8
4.3   Modeling Process	4-10
4.4   Modeling of CAA Compliance Flexibilities	4-13
4.5   Per Vehicle Costs 2012-2016	4-16
4.6   Technology Penetration	4-19
4.7   Manufacturer-Specific Standards	24
4.8   Alternative Program Stringencies	25
4.9   Assessment of Manufacturer Differences	4-33
CHAPTER 5:  EMISSIONS IMPACTS	5-1
5.1   Overview	5-1

-------
                                                Draft Regulatory Impact Analysis
5.2   Introduction	5-3
  5.2.1    Scope of Analysis	5-3
  5.2.2    Downstream Contributions	5-4
  5.2.3    Upstream Contributions	5-4
  5.2.4    Global Warming Potentials	5-5
5.3   Program Analysis and Modeling Methods	5-5
  5.3.1    Models Used	5-5
  5.3.2    Description of Scenarios	5-6
  5.3.3    Calculation of Downstream Emissions	5-14
  5.3.4    Calculation of Upstream Emissions	5-24
5.4   Greenhouse Gas Emission Inventory	5-24
5.5   Non-Greenhouse Gas Emission Inventory	5-26
  5.5.1    Downstream Impacts of Program	5-28
  5.5.2    Upstream Impacts of Program	5-29
  5.5.3    Total Program Impact	5-30
5.6   Model Year Lifetime Analyses	5-31
  5.6.1    Methodology	5-31
  5.6.2    Results	5-33
5.7   Alternative 4% and 6% Scenarios	5-34
  5.7.1    4% Scenario	5-35
  5.7.2    6% Scenario	5-37
5.A  Appendix to Chapter 5:  Details of the TLAAS Impacts Analysis	5-40
  5.A.1   Introduction and Summary	5-40
  5.A.2   Factors Determining the Impact of the TLAAS	5-40
  5.A.3   Bounding Analysis of TLAAS Impact	5-41
  5.A.4   Approach used for Estimating TLAAS Impact	5-42
CHAPTER 6:  VEHICLE PROGRAM COSTS INCLUDING FUEL CONSUMPTION
IMPACTS     6-1
6.1   Vehicle Program Costs	6-1
  6.1.1    Vehicle Compliance Costs on a Per-Vehicle Basis	6-1
  6.1.2    Vehicle Compliance Costs on a Per-Year Basis	6-7
6.2   Cost per Ton of Emissions Reduced	6-11
6.3   Fuel Consumption Impacts	6-12

                                         iii

-------
Draft Regulatory Impact Analysis
6.4   Vehicle Program Cost Summary	6-15
CHAPTER 7:   ENVIRONMENTAL AND HEALTH IMPACTS	7-1
7.1   Health and Environmental Effects of Non-GHG Pollutants	7-1
  7.1.1    Health Effects Associated with Exposure to Pollutants	7-1
  7.1.2    Environmental Effects Associated with Exposure to Pollutants	7-13
7.2   Non-GHG Air Quality Impacts	7-26
  7.2.1    Introduction	7-26
  7.2.2    Current Levels of Pollutants	7-26
  7.2.3    Impacts on Future Air Quality	7-28
7.3   Quantified and Monetized Co-Pollutant Health and Environmental Impacts ....7-31
  7.3.1    Economic Value of Reductions in Criteria Pollutants	7-32
  7.3.2    Human Health and Environmental Benefits for the Final Rule	7-37
7.4   Changes in Global Mean Temperature and Sea-Level Rise Associated with the
Proposal's GHG Emissions Reductions	7-43
  7.4.1    Introduction	7-43
  7.4.2    Estimated Projected Reductions in Global Mean Surface Temperature and
  Sea-Level Rise	7-43
7.5   SCC and GHG Benefits	7-46
7.6   Weight Reduction and Vehicle Safety	7-56
CHAPTER 8:   OTHER ECONOMIC AND SOCIAL IMPACTS	8-1
8.1   Vehicle Sales Impacts	8-1
  8.1.1    How Vehicle Sales Impacts were Estimated for this Rule	8-1
  8.1.2    Consumer Vehicle Choice Modeling	8-4
  8.1.3    Consumer Payback Period and Lifetime Savings on New Vehicle Purchases
          8-11
8.2	8-14
8.3   Energy Security Impacts	8-14
8.4   Other Externalities	8-16
  8.4.1    Reduced Refueling Time	8-16
  8.4.2    Value of Additional Driving	8-17
  8.4.3    Noise, Congestion, and Accidents	8-17
  8.4.4    Summary of Other Externalities	8-18
8.5   Summary of Costs and Benefits	8-20

-------
                                           Draft Regulatory Impact Analysis
CHAPTER 9:  SMALL BUSINESS FLEXIBILITY ANALYSIS	9-1

-------
Draft Regulatory Impact Analysis
List of Acronyms
      2-mode: 2-mode hybrid electric vehicle
      2V:    2-valves per cylinder
      4V:    4-valves per cylinder
      12V:   12 Volts
      42V:   42 Volts
      A/C:   Air conditioner/conditioning
      AERO: Improved aerodynamics
      ASL:   Aggressive Shift Logic
      AT:    Automatic transmission
      CAFE: Corporate Average Fuel Economy
      CCP:   Couple Cam Phasing
      CO2:   carbon dioxide
      CVA:  Camless Valve Actuation (full)
      CVT:  Continuously Variable Transmission
      CVVL: Continuous Variable Valve Lift
      Deac:  Cylinder Deactivation
      DICE: Dynamic Integrated Model of Climate and the Economy
      DCP:   Dual (independent) Cam Phasing
      DCT:  6-speed Dual Clutch Transmission
      DOHC: Dual Overhead Camshafts
      DOT:  Department of Transportation
      DVVL: Discrete (two-step) Variable Valve Lift
      EFR:   Engine Friction Reduction
      EIS:   Environmental Impact Statement
      EPS:   Electric Power Steering
      FUND:Climate Framework for Uncertainty, Negotiation, and Distribution
      GDI:   Gasoline Direct Injection
      GHG:  Greenhouse gas
      HCCI: Homogenous Charge Compression Ignition (gasoline)
      HEV:  Hybrid Electric Vehicle
      13:     In-line 3-cylinder engine
      14:     In-line 4-cylinder engine
      IACC: Improved Accessories
      IAM:   Integrated Assessment Model
      IMA:   Integrated Motor Assist
      IPCC:  Intergovernmental Panel on Climate Change
      L4:    Lock-up 4-speed automatic transmission
      L5:    Lock-up 5-speed automatic transmission
      L6:    Lock-up 6-speed automatic transmission
      LDB:  Low drag brakes
      LRR:   Low Rolling Resistance
      LUB:  Low-friction engine lubricants
      MPV:  Multi-Purpose Vehicle
      MY:   Model Year
                                          VI

-------
                                           Draft Regulatory Impact Analysis
NESCCAF: Northeast States Center for a Clean Air Future
NHTSA: National Highway Transportation Safety Administration
OECD:Organization for Economic Cooperation and Development
OHV:  Overhead Valve (pushrod)
OMB: Office of Management and Budget
ORNL: Oak Ridge National Laboratory
PAGE: Policy Analysis for the Greenhouse Effect
PHEV: Plug-in Hybrid Electric Vehicle
PRTP: Pure Rate of Time Preference
S&P:  Standard and Poor's
SCC:  Social Cost of Carbon
SCR:  Selective Catalytic Reduction
SOHC: Single Overhead Camshaft
SRES: Special Report on Emissions Scenarios
S-S:   Stop-start hybrid system
THC:  Thermohaline circulation
TORQ: Early torque converter lockup
Turbo: Turbocharger/Turbocharging
V6:   6-cyUnder engine in a "V" configuration
V8:   8-cyUnder engine in a "V" configuration
WGII: Working group II
                                    vn

-------
Draft Regulatory Impact Analysis
                              This Page Intentionally Left Blank
                                             Vlll

-------
                                                   Draft Regulatory Impact Analysis
Executive Summary

       The Environmental Protection Agency (EPA) and the National Highway Traffic
Safety Administration (NHTSA) are issuing a joint proposal to establish new standards for
light-duty highway vehicles that will reduce greenhouse gas emissions and improve fuel
economy. The joint proposed rulemaking 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 proposing
greenhouse gas emissions standards under the Clean Air Act, and NHTSA is proposing
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. They would require these
vehicles to meet an estimated combined average emissions level of 250 grams of CCh per
mile in MY 2016 under EPA's GHG program, and 34.1 mpg in MY 2016 under NHTSA's
CAFE program and represent a harmonized and consistent national program (National
Program). These standards are designed such that compliance can be achieved with a single
national vehicle fleet whose emissions and fuel economy performance improves year over
year.  The proposed National Program would result in approximately 950 million metric tons
of CO2 emission reductions and approximately 1.8 billion barrels of oil savings over the
lifetime of vehicles sold in model years 2012 through 2016.

       The vehicle categories covered by the rulemaking are responsible for almost 60
percent of all U.S. transportation-related greenhouse gas emissions and include cars, sport
utility vehicles, minivans, and pickup trucks used for personal transportation.  Transportation
related emissions are responsible for approximately 30 percent of total U.S. greenhouse gas
emissions. Under the National Program, automobile manufacturers would 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.

       This draft regulatory impact analysis (DRIA) contains supporting documentation to
the EPA proposal. NHTSA has prepared their own Proposed RIA (PRIA) in support of their
proposal (this can be found in NHTSA's docket for their proposal, NHTSA-2009-0059).
While the two proposals are similar, there are also differences in the analyses that require
separate discussion. This is largely because EPA and NHTSA act under different statutes.
EPA's authority comes under the Clean Air Act, and NHTSA's authority comes under EPCA,
and each statute has somewhat different requirements and flexibilities. As a result, each
agency has followed a unique approach where warranted by these differences. Where each
agency has followed the same approach—e.g., development of technology costs and
effectiveness—the supporting documentation is contained in the draft joint technical support
document (draft joint TSD, which can be found in EPA's docket EPA-HQ-OAR-2009-0472).
Therefore, this DRIA should be viewed as a companion document to the draft joint TSD and
the two documents together provide the details of EPA's technical analysis in support of its
proposal.
                                            IX

-------
Draft Regulatory Impact Analysis
       Specifically, this document contains, in Chapter 1, a description of EPA's use of
technology packages in the OMEGA model. This discussion builds on the discussion
contained in Chapter 3 of the draft joint TSD which provides details of technology costs and
effectiveness but only an overview of how technologies are put together into packages for the
OMEGA model. Chapter 1 also contains a discussion of the lumped parameter model which
is a major part of our determination of the effectiveness of these packages.

       In Chapter 2, we present a detailed discussion of our AC credit program and the
technology costs and effectiveness associated with new AC systems. This discussion is
unique to this DRIA as the AC-related proposal is unique to EPA.

       In Chapter 3, we present the technical basis of EPA's proposed standards and an
analysis of the "footprint" approach EPA is proposing for establishing standards.  In Chapter
4, we present an overview of the OMEGA model and the modeling results (actual OMEGA
model inputs and outputs) in support of the proposed program and the alternative standards
that were considered. Chapter 5 presents the emission reductions expected from the  proposal.
Chapter 6 presents the program costs and fuel savings associated with EPA's proposal.
Chapter 7 presents the environmental and health impacts, including EPA's discussion of the
social cost of carbon, and Chapter 8 presents other economic and social impacts—e.g., less
time spent refueling due to higher fuel efficiency—of the proposal.  Chapter 9 presents our
analysis of the small business impacts due to EPA's proposal. All of these discussions—
Chapters 3 through 9—are unique to this DRIA since, even though many of the metrics are
common between EPA and DOT, we have different results due to our use of different models
(EPA's OMEGA model versus DOT's CAFE Compliance and Effects Modeling System
(often referred to as "the CAFE model" or "the Volpe model")) and the differences in our
programs (e.g., AC credits versus no AC credits, plus many other program flexibilities).

Greenhouse Emission Impacts of EPA's  Proposal

       Table 1 shows reductions estimated from EPA's proposed GHG standards assuming a
pre-control case of 2011 MY CAFE standards continuing indefinitely beyond 2011, and a
post-control case in which 2016  MY standards continue indefinitely beyond 2016. These
reductions are broken down by upstream and downstream components, including air
conditioning improvements, and also account for the offset from a 10 percent "rebound"
effect in vehicle miles travelled as discussed in Chapter 4 of the joint draft TSD.A Including
the reductions from upstream emissions, total reductions are estimated to reach 325
MMTCO2eq (million metric tons of CO2 equivalent emissions) annually by 2030 (a 21
percent reduction in U.S. car and light truck  emissions), and grow to over 500 MMTCO2eq in
2050 as cleaner vehicles continue to come into the fleet (a 23 percent reduction in U.S. car
and light truck emissions).
A "Rebound VMT" is the term used to describe the increase in driving that might occur as vehicle fuel
consumption decreases (i.e., the fuel economy improves) since the cost per mile of operating the vehicle
decreases. As a result of this rebound effect, the benefits of the proposed rule are offset slightly since owners of
compliant vehicles drive more miles resulting in slightly more GHG emissions.  Importantly, the adverse effects,
or disbenefits, of rebound VMT are far outweighed by the overall benefits of the proposal.

-------
                                                    Draft Regulatory Impact Analysis
             Table 1. Projected Net GHG Reductions (MMT CO2 Equivalent per year)
CALENDAR YEAR
Net reduction to
tailpipe standards*
Tailpipe standards
A/C -indirect CO 2
A/C- direct HFCs
Upstream

Percent reduction
relative to U.S.
reference (cars + light
trucks)
Percent reduction
relative to U.S.
reference (all sectors)
Percent reduction
relative to worldwide
reference
2020
165.2
107.7
11.0
13.5
33.1

12.4%
2.2%
0.3%
2030
324.6
211.4
21.1
27.2
64.9

2.14%
4.2%
0.6%
2040
417.5
274.1
27.3
32.1
84.1

22.8%
5.2%
0.7%
2050
518.5
344.0
34.2
34.9
105.5

22.9%
6.2%
0.9%
            : includes impacts of 10% VMT rebound rate presented in Table III.F.I-3
Criteria Pollutant Impacts of EPA's Proposal

       As shown in Table 2, EPA estimates that the proposed program would result in
reductions of oxides of nitrogen (NOx), volatile organic compounds (VOC), paniculate matter
(PM) and oxides of sulfur (SOx), but would increase carbon monoxide (CO) emissions.  The
CO increase is because gasoline fueled passenger cars and light trucks contribute over 50
percent of the total CO emissions in the US, whereas for other pollutants the contribution is
less than 40 percent. Thus, for CO the increase from VMT rebound outweighs the upstream
CO reductions.  For all pollutants the overall impact of the program would be relatively small
compared to total U.S. inventories across all sectors.  In the year 2030, EPA estimates the
proposed program would reduce these total NOx, PM and SOx inventories by  0.2 to 0.3
percent and reduce the VOC inventory by 1.2 percent, while increasing the total national CO
inventory by 0.4 percent.

       As shown in Table 3, EPA estimates that the proposed program would result in small
changes for toxic emissions compared to total U.S. inventories across all sectors. In 2030,
EPA estimates the program would reduce total benzene and formaldehyde by 0.04 percent.
Total acrolein, acetaldehyde, and 1,3-butadiene would increase by 0.03 to 0.2 percent.
                                            XI

-------
Draft Regulatory Impact Analysis
       Other factors which may impact criteria, or non-GHG, emissions but are not estimated
in this analysis include:

          •  Vehicle technologies used to reduce tailpipe CCh emissions; because the
             regulatory standards for non-GHG emissions are the primary driver for these
             emissions, EPA expects the impact of today's program to be negligible on non-
             GHG emission rates per mile.

          •  The potential for increased market penetration of diesel vehicles; because
             these vehicles would be held to the same certification and in-use standards for
             criteria pollutants as their gasoline counterparts, EPA expects their impact to
             be negligible on criteria pollutants and other non-GHG emissions.

          •  Early introduction of electric vehicles and plug-in hybrid electric vehicles,
             which would reduce criteria emissions in cases where they are able to certify to
             lower certification standards. It would also likely reduce gaseous air toxics.

          •  Reduced refueling emissions due to less frequent refueling events and reduced
             annual refueling volumes resulting from the GHG standards.

          •  Increased hot soak evaporative  emissions due to the likely increase in number
             of trips associated with VMT rebound modeled in this proposal.

          •  Increased market share of E10 relative to  EO, due to the decreased overall
             gasoline consumption of today's proposal combined with an unchanged fuel
             ethanol volume.

                Table 2. Annual Criteria Emission Impacts of Program (short tons)

voc
% of total inventory
CO
% of total inventory
NOX
% of total inventory
PM2.5
% of total inventory
SOx
% of total inventory
TOTAL IMPACTS
2020
-73,739
-0.60%
70,614
0.13%
-17,206
-0.14%
-2,856
-0.08%
-16,307
-0.18%
2030
-142,347
-1.20%
227,832
0.38%
-27,726
-0.20%
-5,431
-0.16%
-31,965
-0.34%
UPSTREAM
IMPACATS
2020
-75,437
-0.61%
-7,209
-0.01%
-22,560
-0.18%
-3,075
-0.09%
-13,804
-0.16%
2030
-147,841
-1.20%
-14,107
-0.02%
-43,286
-0.36%
-6,003
-0.18%
-27,060
-0.29%
DOWNSTREAM
IMPACTS
2020
1,698
0.01%
77,823
0.14%
5,354
0.04%
218
0.01%
-2,503
-0.03%
2030
5,494
0.05%
241,939
0.40%
15,560
0.13%
572
0.02%
-4,906
-0.05%
                                             Xll

-------
                                                    Draft Regulatory Impact Analysis
               Table 3. Annual Air Toxic Emission Impacts of Program (short tons)

1,3-Butadiene
% of total inventory
Acetaldehyde
% of total inventory
Acrolein
% of total inventory
Benzene
% of total inventory
Formaldehyde
% of total inventory
TOTAL IMPACTS
2020
11
0.07%
17
0.04%
0
0.00%
-84
-0.04%
-28
-0.03%
2030
37
0.22%
61
0.13%
2
0.03%
-77
-0.04%
-16
-0.02%
UPSTREAM
IMPACATS
2020
-1.8
-0.01%
-8
-0.02%
-1.1
0.00%
-163
-0.08%
-60
-0.07%
2030
-3.4
-0.02%
-15
-0.03%
-2
0.00%
-320
-0.15%
-112
-0.10%
DOWNSTREAM
IMPACTS
2020
13.2
0.08%
24.8
0.05%
1.3
0.02%
79.6
0.04%
31.8
0.04%
2030
40.2
0.24%
75.5
0.17%
3.9
0.06%
242.2
0.11%
96.3
0.11%
Costs and Benefits of EPA's Proposal

       Table 4 presents estimated annual net benefits for the indicated calendar years.  The
table also shows the net present values of those net benefits for the calendar years 2012-2050
using both a 3 percent and a 7 percent discount rate.  The table includes the benefits of
reduced GHG emissions—and consequently the annual net benefits—for each of five interim
SCC values considered by EPA (please refer to Chapter 7 of this DRIA for a discussion of the
five interim SCC values). As noted in Chapter 7, there is a very high probability (very likely
according to the IPCC) that the benefit estimates from GHG reductions are underestimates
because, in part, models used to calculate SCC values do not include information about
impacts that have not been quantified. Note that the  quantified annual costs shown in Table 4
are negative because fuel savings are included. Fuel savings are  considered as negative costs
(i.e., positive benefits) of the proposed vehicle program.  The fuel savings outweigh the costs
associated with the addition of new technology and, therefore,  the vehicle program costs are
negative.
                                            Xlll

-------
Draft Regulatory Impact Analysis
   Table 4 Quantified Net Benefits Associated with the Proposed Light-Duty Vehicle GHG Program3'b

                                       (Millions of 2007 dollars)

Quantified Annual
Costs
2020
-$25,100
2030
-$72,500
2040
-$105,700
2050
-$146,100
NPV, 3%
-$1,287,600
NPV, 7%
-$529,500
Quantified Annual Benefits at each assumed SCC value
SCC 5%
SCC 5% Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC 3% Newell-Pizer
$9,900
$11,200
$13,400
$16,900
$22,700
$21,100
$24,400
$29,800
$39,800
$53,800
$30,200
$35,500
$46,500
$62,500
$87,500
$42,100
$51,600
$68,600
$95,600
$132,600
$400,900
$470,100
$594,700
$788,600
$1,093,100
$177,200
$205,700
$257,100
$337,100
$462,800
Quantified Net Benefits at each assumed SCC value
SCC 5%
SCC 5% Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC 3% Newell-Pizer
$35,000
$36,300
$38,500
$42,000
$47,800
$93,600
$96,900
$102,300
$112,300
$126,300
$135,900
$141,200
$152,200
$168,200
$193,200
$188,200
$197,700
$214,700
$241,700
$278,700
$1,688,500
$1,757,700
$1,882,300
$2,076,200
$2,380,700
$706,700
$735,200
$786,600
$866,600
$992,300
"Note that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts. Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only human
health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental co-
benefits would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality
modeling. However, we were unable to conduct a full-scale air quality modeling analysis in time for the
proposal. We intend to more fully capture the co-pollutant benefits for the analysis of the final standards.
b Quantified annual costs are shown as negative here because fuel savings are included.  Fuel savings are
considered as negative costs (i.e., positive benefits) of the proposed vehicle program.  The fuel savings outweigh
the costs associated with the addition of new technology and, therefore, the vehicle program costs are negative.
The fuel impacts included here were calculated using pre-tax fuel prices.
                                                     XIV

-------
                                        Technology Packages, Cost and Effectiveness
CHAPTER 1: Technology Packages, Cost and Effectiveness
1.1 Overview of Technology

       The proposed GHG program 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 in the 2012-2016 time frame.  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 (both per manufacturer and per
vehicle) and  other impacts of implementing the technology, and the lead time needed for
manufacturers to employ the control technology. EPA also considers 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 and unquantified benefits, safety, and other impacts.
The availability of technology to achieve reductions and the cost and other aspects  of this
technology are therefore a central focus of this rulemaking.

       At the same time, the technological problems and solutions involved in this
rulemaking differ in many ways from prior mobile source rulemakings. In the past the
assessment of exhaust emissions control technology has focused on how to reduce the amount
of various unwanted chemical compounds that are generated when fuel is combusted.  The
emissions are often the result of incomplete combustion, such as emissions of HC, CO, and
PM.  In some cases the combustion products  are the result of the specific conditions under
which combustion occurs, such as the relationship between emissions of NOx and the
temperature of combustion. Technology to control exhaust emissions has focused, in part,  on
changing the fuel delivery and engine systems so there is more complete combustion of the
fuel which generates less HC, CO, and PM in the engine exhaust but, by design, generates
more CO2. (CO2 is one of ultimate  combustion products of any carbon containing fuel, such
as gasoline and diesel fuel.)  Other changes to the fuel delivery and engine systems have
been designed to change the combustion process to reduce the amount of NOx and PM
generated by the engine. Very large reductions have been achieved by installing and
optimizing aftertreatment (post-combustion, post- engine generated pollution) devices, such
as catalytic converters and catalyzed diesel paniculate filters (DPF), that reduce the amount of
emissions of HC, CO, and PM by oxidizing or combusting these compounds in the
aftertreatment device, again  generating CO2  in the process.  In the case of NOx,
aftertreatment devices have focused on the chemical process of reduction, or removal of
oxygen from the compound. Therefore the exhaust emissions control technologies of the past
have focused almost exclusively on (1) upgrading the fuel delivery and engine systems to
control the combustion process to reduce the  amount of unwanted emissions from the engine
and in the process increase the amount of CO2 emitted, and on (2) aftertreatment devices that
either continue this oxidation process and increase emissions of CO2, or otherwise change the
compounds emitted by the engine. Since CO2 is a stable compound produced by the complete
combustion of the fuel - indeed serving as a marker of how efficiently fuel has been
combusted, these two methods employed to address HC, CO, PM, and NOx are not available
                                           1-1

-------
Draft Regulatory Impact Analysis
to address €62. Instead, the focus of the CO2 emissions control technology must be entirely
different—reducing the amount of fuel that is combusted.

       Vehicles combust fuel to perform two basic functions: 1) transport the vehicle, its
passengers and its contents, and 2) operate various accessories during the operation of the
vehicle such as the air conditioner.  Technology can reduce €62 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 involves a major change in focus and calls for looking at the
vehicle as an entire system. 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.

       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.

       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 proposed standards.  Vehicle manufacturers
typically develop their many different models by basing them on a limited number of vehicle
platforms. Several different models of vehicles are produced using a common platform,
allowing for efficient use of design and manufacturing resources. The platform typically
consists of common vehicle architecture and structural components.  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 all of the other changes
needed so the vehicle model will meet the manufacturer's plans for the next several years.
This includes meeting all of the emissions and other requirements that would apply during the
years before the next major redesign of the vehicle.

       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

                                             1-2

-------
                                          Technology Packages, Cost and Effectiveness
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., aerodynamic improvements, valve timing improvements).
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.

       As discussed below, there are a wide variety of emissions  control 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 heads, or
redesign of the transmission and its packaging in the vehicle. This calls for tying the
incorporation of the emissions control technology into 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.

       Over the five model  years at issue in this rulemaking, 2012-2016, EPA projects that
almost the entire fleet of light-duty vehicles (i.e., 85 percent) will have gone through a
redesign cycle.  If the technology to control greenhouse gas emissions is efficiently folded
into this redesign process, then by 2016  almost the entire light-duty fleet could be designed to
employ upgraded packages  of technology to reduce emissions of CCh, and as discussed
below, to reduce emissions of HFCs from the air conditioner.

       In determining the requisite technology and cost of these first ever GHG emissions
standards for light-duty vehicles, EPA proposes to use 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 the GHG proposal is
commercially available and already being employed to a limited extent across the fleet.  The
vast majority of the emission reductions which would result from the proposed rule would
result from the increased the use of these technologies. EPA also believes the proposed rule
would encourage the development and limited use of more advanced technologies, such as
PHEVs and EVs.

       In section 1.2 below, a summary of technology costs and effectiveness is presented. In
section 1.3, the process of combining technologies into packages is described along with
package costs and effectiveness.  Sections 1.4 through 1.6 discuss the lumped parameter
approach which provides background and support for determining technology and package
effectiveness.

                                             1-3

-------
Draft Regulatory Impact Analysis
1.2 Technology Cost and Effectiveness

       EPA collected information on the cost and effectiveness of CO2 emission reducing
technologies from a wide range of sources.  The primary sources of information were
NHTSA's 2011 CAFE FRM and EPA's 2008 Staff Technical Report.  In those analyses,
piece costs and effectiveness were estimated based on a number of sources.  The objective
was to use those sources of information considered to be most credible. Those sources
included: the 2002 NAS report on the effectiveness and impact of CAFE standards; the 2004
study done by the Northeast States Center for a Clean Air Future (NESCCAF); the
California Air Resources Board (CARB) Initial Statement of Reasons in support of their
carbon rulemaking; a 2006 study done by Energy and Environmental Analysis (EEA) for the
Department of Energy;  a study done by the Martec Group for the Alliance of Automobile
Manufacturers, and an update by the Martec Group to that study; and vehicle fuel economy
certification data.  In addition, confidential data submitted by vehicle manufacturers in
response to NHTSA's request for product plans were considered, as was confidential
information shared by automotive industry component suppliers in meetings with EPA and
NHTS A staff held during the second half of the 2007 calendar year. These confidential data
sources were used primarily as a validation of the estimates since EPA prefers to rely on
public data rather than confidential data. EPA also has a contracted study ongoing with FEV
(an engineering services firm) that consists of complete system tear-downs to evaluate
technologies down to the nuts and bolts to arrive  at very detailed estimates of the costs
associated with manufacturing them.  Lastly, cost and effectiveness estimates have been
adjusted slightly as a result of further meetings between EPA and NHTS A staff in the  first
half of 2009 where both piece costs and fuel consumption efficiencies were discussed  in
detail.  EPA also reviewed the published technical literature which addressed the issue of CCh
emission control, such as papers published by the Society of Automotive Engineers and the
American Society of Mechanical Engineers. The results of all of the research and discussions
are summarized in Chapter 3 of the draft Joint Technical Support Document.

       EPA reviewed all this information in order to develop the best estimates of the cost
and effectiveness of CCh reducing technologies.  These estimates were developed for five
vehicle classes:  small car, large car, minivan,  small truck and large truck. All vehicle types
were mapped into one of these five classes in EPA's  analysis (see Chapter 3 of the draft Joint
TSD).  Fuel consumption reductions are possible from a variety of technologies whether they
be engine-related (e.g., turbocharging), transmission-related (e.g., six forward gears in place
of four), accessory-related  (e.g., electronic power steering), or vehicle-related (e.g., low
rolling resistance tires).  Table 1-1 through Table 1-5 show estimates of the near term cost
associated with various technologies for the five vehicle classes used in this analysis.  These
estimates shown in Table 1-1 through Table 1-5 are relative to a baseline vehicle having a
multi-point, port fuel injected gasoline engine operating at a stoichiometric air-fuel ratio with
fixed valve timing and lift and without any turbo  or super charging and equipped with a 4-
speed automatic transmission. This configuration was chosen as the baseline vehicle because
it is the predominant technology package sold in  the  United States. Costs are presented in
terms of their hardware incremental compliance cost. This means that they include all
potential costs associated with their application on vehicles, not just the cost of their physical
parts. A more detailed description of these and the following estimates of cost and
                                             1-4

-------
                                           Technology Packages, Cost and Effectiveness
effectiveness of CC>2 reducing technologies can be found in Chapter 3 of the draft joint TSD,
along with a more detailed description of the comprehensive technical evaluation underlying
the estimates.

 Table 1-1 EPA's Incremental Piece Costs for Engine Technologies Marked up to include both Direct and
                        Indirect Costs in 2016 (2007 Dollars per Vehicle)
Technology

OHC Engines
OHV Engines

Turbo w/o downsize
Downsize w/o turbo
Turbo with downsize
Low friction lubricants
Engine friction reduction
WT - intake cam
phasing
WT - coupled cam
phasing
WT - dual cam phasing
Cylinder deactivation
Discrete VVLT
Continuous VVLT
Cylinder deactivation
VVT — coupled cam
phasing
Discrete VVLT
Continuous WLT
(includes conversion to
Overhead Cam)
Camless valvetrain
(electromagnetic)
GDI - stoichiometric
GDI — lean bum
Turbocharge (single)
Turbocharge (twin)
Downsize to 14 DOHC
Downsize to 14 DOHC
Downsize to 14 DOHC
Downsize to 14 DOHC
Downsize to 13 DOHC
Downsize to V6 DOHC
Downsize to V6 DOHC
Downsize to V6 DOHC
Downsize to V6 DOHC
Downsize to 14 DOHC
& add turbo
Downsize to 14 DOHC
& add turbo
Downsize to 14 DOHC
& add turbo
Downsize to 14 DOHC
& add turbo
Downsize to 13 DOHC
& add turbo
Downsize to V6 DOHC
& add twin turbo
Downsize to V6 DOHC
& add twin turbo
Downsize to V6 DOHC
& add twin turbo
Downsize to V6 DOHC
Incremental to
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine
Base engine w/ VVT-
coupled
Base engine
Base engine
GDI - stoich
Base engine
Base engine
V6 DOHC
V6 SOHC
V6OHV
14 DOHC (larger)
14 DOHC
V8 DOHC
V8 SOHC 2V
V8 SOHC 3V
V8OHV
V6 DOHC w/o turbo
V6 SOHC w/o turbo
V6 OHV w/o turbo
14 DOHC (larger)
w/o turbo
14 DOHC w/o turbo
V8 DOHC w/o turbo
V8 SOHC 2V w/o
turbo
V8 SOHC 3V w/o
turbo
V8 OHV w/o turbo
Vehicle Class
Small Car
$3
$50
$40
$40
$73
n/a
$125
$245
n/a
$40
$141
$497
$501
$222
$623
$366
$663
-$337
-$53
$265
-$47
-$80
n/a
n/a
n/a
n/a
$214
$453
$797
$372
$344
n/a
n/a
n/a
n/a
Large Car
$3
$75
$80
$80
$157
$150
$181
$449
$150
$40
$204
$1,048
$501
$287
$623
$366
$663
-$337
-$53
$265
-$47
n/a
-$160
$199
$111
$310
$214
$453
$797
$372
n/a
$613
$971
$872
$1,096
Minivan
$3
$75
$80
$80
$157
$150
$181
$449
$150
$40
$204
$1,048
$501
$287
$623
$366
$663
-$337
-$53
$265
-$47
n/a
-$160
$199
$111
$310
$214
$453
$797
$372
n/a
$613
$971
$872
$1,096
Small Truck
$3
$75
$80
$80
$157
$150
$181
$449
$150
$40
$204
$1,048
$501
$287
$623
$366
$663
-$337
-$53
$265
-$47
n/a
-$160
$199
$111
$310
$214
$453
$797
$372
n/a
$613
$971
$872
$1,096
Large Truck
$3
$100
$80
$80
$157
$169
$259
$489
$169
$40
$291
$1,146
$501
$312
$623
$366
$663
-$337
-$53
$265
-$47
n/a
-$160
$199
$111
$310
$214
$453
$797
$372
n/a
$613
$971
$872
$1,096
                                              1-5

-------
Draft Regulatory Impact Analysis



& add twin turbo
Convert to V6 DOHC
Convert to V6 DOHC
Convert to V8 DOHC
Convert to V8 DOHC
Convert to V8 DOHC
Gasoline HCCI dual-
mode
Diesel - Lean NOx trap
Diesel - urea SCR

V6 SOHC
V6OHV
V8 SOHC 2V
V8 SOHC 3V
V8OHV
GDI - stoich
Base gasoline engine
Base gasoline engine

n/a
n/a
n/a
n/a
n/a
$253



$354
$464
$398
$310
$509
$375

$2,655

$354
$464
$398
$310
$509
$375

$2,164

$354
$464
$398
$310
$509
$375

$2,164

$354
$464
$398
$310
$509
$659

$2,961
   Table 1-2 EPA's Incremental Piece Costs for Transmission Technologies Marked up to include both
                    Direct and Indirect Costs in 2016 (2007 Dollars per Vehicle)
Technology
Aggressive shift logic
Early torque converter lockup
5-speed automatic
6-speed automatic
6-speed DCT - dry clutch
6-speed DCT - wet clutch
6-speed manual
CVT
Incremental to
Base trans
Base trans
4-speed auto trans
4-speed auto trans
6-speed auto trans
6-speed auto trans
5-speed manual trans
4-speed auto trans
Vehicle Class
Small
Car
$28
$25
$90
$150
$65
$139
$79
$192
Large
Car
$28
$25
$90
$150
$65
$139
$79
$224
Minivan
$28
$25
$90
$150
$65
$139
$79
$224
Small
Truck
$28
$25
$90
$150
$65
$139
$79
n/a
Large
Truck
$28
$25
$90
$150
$65
$139
$79
n/a
 Table 1-3 EPA's Incremental Piece Costs for Hybrid Technologies Marked up to include both Direct and
                         Indirect Costs in 2016 (2007 Dollars per Vehicle)
Technology
Stop- Start
IMA/ISA/BSG
(includes engine
downsize)
2-Mode hybrid
electric vehicle
Power-split
hybrid electric
vehicle
Full-Series
hydraulic hybrid
Plug-in hybrid
electric vehicle
Plug-in hybrid
electric vehicle
Full electric
vehicle
Incremental to
Base engine &
trans
Base engine &
trans
Base engine &
trans
Base engine &
trans
Base engine &
trans
IMA/ISA/BSG
hybrid
Power-split
hybrid
Base engine &
trans
Vehicle Class
Small Car
$351
$2,854
$4,232
$3,967
$750
$6,922
$5,423
$27,628
Large Car
$398
$3,612
$5,469
$5,377
$825
$9,519
$7,431
n/a
Minivan
$398
$3,627
$5,451
$5,378
$825
$9,598
$7,351
n/a
Small Truck
$398
$3,423
$4,943
$4,856
$900
$9,083
$7,128
n/a
Large Truck
$437
$4,431
$7,236
$7,210
$1200
$12,467
$9,643
n/a
                                                 1-6

-------
                                            Technology Packages, Cost and Effectiveness
 Table 1-4 EPA's Incremental Piece Costs for Accessory Technologies Marked up to include both Direct
                      and Indirect Costs in 2016 (2007 Dollars per Vehicle)
Technology
Improved high efficiency
alternator &
electrification of
accessories
Upgrade to 42 volt
electrical system
Electric power steering
(12 or 42 volt)
Incremental to
Base accessories
12 volt electrical
system
Base power
steering
Vehicle Class
Small Car
$76
$86
$94
Large Car
$76
$86
$94
Minivan
$76
$86
$94
Small Truck
$76
$86
$94
Large Truck
$76
$86
$94
 Table 1-5 EPA's Incremental Piece Costs for Vehicle Technologies Marked up to include both Direct and
                        Indirect Costs in 2016 (2007 Dollars per Vehicle)
Technology
Aero drag reduction (20%
on cars, 10% on trucks)
Low rolling resistance
tires
Low drag brakes (ladder
frame only)
Secondary axle disconnect
(unibody only)
Front axle disconnect
(ladder frame only)
Incremental to
Base vehicle
Base tires
Base brakes
Base vehicle
Base vehicle
Vehicle Class
Small Car
$42
$6
n/a
$514
n/a
Large Car
$42
$6
n/a
$514
n/a
Minivan
$42
$6
n/a
$514
n/a
Small Truck
$42
$6
$63
$514
$84
Large Truck
$42
$6
$63
n/a
$84
       Table 1-6 through Table 1-10 summarize the CCh reduction estimates of various
technologies which can be applied to cars and light-duty trucks. A similar summary of costs
is provided in Chapter 3 of the draft joint TSD and each of these estimates is discussed in
more detail there.
                                               1-7

-------
Draft Regulatory Impact Analysis
                          Table 1-6 Engine Technology Effectiveness
Technology
Low friction lubricants — incremental to base engine
Engine friction reduction - incremental to base engine
Absolute CO2 Reduction (% from baseline vehicle)
Small Car
0.5
1-3
Large
Car
0.5
1-3
Minivan
0.5
1-3
Small
Truck
0.5
1-3
Large
Truck
0.5
1-3
Overhead Cam Branch
WT - intake cam phasing
WT - coupled cam phasing
VVT — dual cam phasing
Cylinder deactivation (includes imp. oil pump,
if applicable)
Discrete VVLT
Continuous VVLT
2
3
3
n.a.
4
5
1
4
4
6
3
6
1
2
2
6
3
4
1
3
2
6
4
5
2
4
4
6
4
5
Overhead Valve Branch
Cylinder deactivation (includes imp. oil
pump, if applicable)
VVT - coupled cam phasing
Discrete VVLT
Continuous VVLT (includes conversion to
Overhead Cam)
n.a.
3
4
5
6
4
4
6
6
2
3
4
6
3
4
5
6
4
4
5

Camless valvetrain (electromagnetic) **
Gasoline Direct Injection— stoichiometric (GDI-S)
Gasoline Direct Injection-lean burn (incremental to
GDI-S) **
Gasoline HCCI dual-mode (incremental to GDI-S) **
Turbo+downsize (incremental to GDI-S)
Diesel - Lean NOx trap[ ]*
Diesel - urea SCR [ ]*
5-15
1-2
8-10
10-12
5-7
15-26
[25-35]
15-26
[25-35]
5-15
1-2
9-12
10-12
5-7
21-32
[30-40]
21-32
[30-40]
5-15
1-2
9-12
10-12
5-7
21-32
[30-40]
21-32
[30-40]
5-15
1-2
9-12
10-12
5-7
21-32
[30-40]
21-32
[30-40]
5-15
1-2
10-14
10-12
5-7
21-32
[30-40]
21-32
[30-40]
       * Note:  estimates for % reduction in fuel consumption are presented in brackets.




       ** Note: for reference only, not used in this rulemaking
                                             1-8

-------
                     Technology Packages, Cost and Effectiveness
Table 1-7 Transmission Technology Effectiveness
Technology
5-speed automatic (from 4-speed auto)
Aggressive shift logic
Early torque converter lockup
6-speed automatic (from 4-speed auto)
6-speed AMT (from 4-speed auto)
6-speed manual (from 5-speed manual)
CVT (from 4-speed auto)
Absolute CO2 Reduction (% from baseline vehicle)
Small
Car
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
6
Large
Car
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
6
Minivan
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
6
Small
Truck
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
n.a.
Large
Truck
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
n.a.
   Table 1-8 Hybrid Technology Effectiveness
Technology
Stop-Start with 42 volt system
IMA/ISA/BSG (includes engine downsize)
2-Mode hybrid electric vehicle
Power-split hybrid electric vehicle
Full-Series hydraulic hybrid
Plug-in hybrid electric vehicle
Full electric vehicle (EV)
Absolute CC>2 Reduction (% from baseline vehicle)
Small
Car
7.5
30
n.a.
35
40
58
100
Large
Car
7.5
25
40
35
40
58
100
Minivan
7.5
20
40
35
40
58
n.a.
Small
Truck
7.5
20
40
35
40
58
n.a.
Large
Truck
7.5
20
25
n.a.
30
47
n.a.
  Table 1-9 Accessory Technology Effectiveness
Technology
Improved high efficiency alternator & electrification of
accessories (12 volt)
Electric power steering (12 or 42 volt)
Improved high efficiency alternator & electrification of
accessories (42 volt)
Absolute CC>2 Reduction (% from baseline vehicle)
Small
Car
1-2
1.5
2-4
Large
Car
1-2
1.5-2
2-4
Minivan
1-2
2
2-4
Small
Truck
1-2
2
2-4
Large
Truck
1-2
2
2-4
Table 1-10 Other Vehicle Technology Effectiveness
Technology
Aero drag reduction (20% on cars, 10% on trucks)
Low rolling resistance tires (10%)
Low drag brakes (ladder frame only)
Secondary axle disconnect (unibody only)
Front axle disconnect (ladder frame only)
Absolute CO2 Reduction (% from baseline vehicle)
Small
Car
3
1-2
n.a.
1
n.a.
Large
Car
3
1-2
n.a.
1
n.a.
Minivan
3
1-2
n.a.
1
n.a.
Small
Truck
2
1-2
1
1
1.5
Large
Truck
2
n.a.
1
n.a.
1.5
                        1-9

-------
Draft Regulatory Impact Analysis
1.3 Package Cost and Effectiveness

1.3.1 Explanation of Technology Packages

       Individual technologies can be used by manufactures to achieve incremental CO2
reductions. However, as mentioned in Section 1.1, EPA believes that manufacturers are more
likely to bundle technologies into "packages" to capture synergistic aspects and reflect
progressively larger CCh reductions with additions or changes to any given package.  In
addition, manufacturers typically apply new technologies in packages during model
redesigns—which occur once roughly 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.

       Therefore, the approach taken here is to group technologies into packages of
increasing cost and effectiveness. EPA determined that 19 different vehicle types provided
adequate resolution required 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, 14, 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 Figure 1-1. While the five classes
provide adequate resolution for the cost basis associated with technology application, they do
not adequately account for all vehicle attributes such as base vehicle powertrain configuration
and mass reduction. For example, costs and effectiveness estimates for the  small car class
were used to represent costs 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.  Note also that these 19 vehicle types span the  range of vehicle footprints—
smaller footprints for smaller vehicles and larger footprints for larger vehicles—which serve
as the basis for the proposed GHG standards.

       Within each of the 19 vehicle types multiple technology packages were created in
increasing technology content and, hence, increasing effectiveness. Important to note is 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 accessories, and  low  drag brakes.  Subsequent  packages include
advanced gasoline engine and transmission technologies such as turbo/downsizing, GDI, mass
reduction and dual-clutch transmission. The most technologically advanced packages within
a segment included HEV, PHEV and EV designs.  The  end result being 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.
                                            1-10

-------
                                          Technology Packages, Cost and Effectiveness
       The final step in creating the vehicle packages was to evaluate each package within
the 19 vehicle types for cost-effectiveness. This was accomplished by dividing the
incremental cost of the technology package by its incremental effectiveness and assessing the
overall step in cost-effectiveness. Technology packages that demonstrated little to no increase
in effectiveness and a significant increase in cost were eliminated as a choice for the model.
This process provided several positive aspects in the package creation:

       (1) Vehicle packages were not limited by any preconceived assumptions of which
          technologies should be more prominent. An example of this is turbo-downsizing a
          V6 engine. In some cases the GDI V6 with advanced valvetrain technology was
          just as effective as a turbo charge  14, thus excluding the additional cost of turbo
          charging;

       (2) The OMEGA model was allowed to apply packages in an increasing order of both
          effectiveness and cost.

        Some of the  intermediate packages were not cost-effective. As a result, the model
might be blocked from choosing a subsequent package that was cost-effective.  Most of the
diesel packages and some of the hybrid packages exhibited this condition. Due to the high
cost of these packages, and effectiveness on par with advanced gas, the model would not
move through these packages and choose a more cost effective package, thus blocking the
model's logical progression.  This is the reason for the absence of diesel and hybrid packages
in some of the 19 vehicle types available for the OMEGA model.  The specific  criteria used to
remove certain packages from use the model inputs is discussed further below.  It is important
to note that the burning of diesel fuel generates approximately 15% more CO2 than gasoline.
As this rule is based on the reduction of CO2 emissions and not on fuel economy, this creates
an additional effectiveness disadvantage for the diesel packages as compared to the advanced
gas and gas hybrid packages.
                      Figure 1-1 Scaling classes to Vehicle Type Mapping
                                            1-11

-------
Draft Regulatory Impact Analysis
1.3.2 Technology Package Costs & Effectiveness

       As described above, technology packages were created for each of 19 different vehicle
types. These packages are described in Table 1-11 and the 2016 MY costs for each package
are also presented.  Note that Table 1-11 includes all the packages created and considered by
EPA. Only a subset of these packages was actually used as inputs to the OMEGA model
because some of the packages were not desirable from a cost effectiveness standpoint (in
other words, some packages would be skipped over if the next package provide more
attractive cost effectiveness).  Table 1-12 shows the package costs for the packages that were
actually used as inputs to the OMEGA model (note that all packages are listed but only those
for which costs are shown were actually used in the OMEGA model).  This table shows the
package costs for each model year 2012 through 2022 and later.  This shows the impact of
both learning effects and short-term versus long-term indirect cost markups on the package
costs. For details of the learning effects and indirect cost markups used in this analysis refer
to Chapter 3 of the  draft joint TSD. By taking a simple average of the technology package
costs for each year  shown in Table 1-12 and then normalizing the averages to the 2016  model
year average, the package costs for each year can be expressed as a percentage relative  to
2016. These results are shown in Table 1-13.  This table shows that package costs are,  on
average, 117% of the costs for 2016.  This higher cost is due to backing out the learning
effects that are built into the 2016 model year estimates. For 2014, the costs are 108% of
those for 2016 as learning has occurred between 2012 and 2014. The costs for 2022 are 92%
of those for 2016.  This is the result of the long-term ICM kicking in as some indirect costs
are no longer attributable to the proposed program. Table 1-12 also shows the effectiveness
of each package used in the OMEGA model (note that the effectiveness of packages does not
change with model year).
                                           1-12

-------
                                                                                    Technology Packages, Cost and Effectiveness
Table 1-11 Package Descriptions and 2016MY Costs for 19 Vehicle Types (T1-T19), All Packages Considered, Costs in 2007 dollars






Vehicle

Q_
0 ^
13 "*
w ~


,_
o „
« b-
|-^r
o



Q_
^
1 s;
,£ -o co1
™ £ t-
31-
Q) ^
.N
^


« _
W -o"
^ -Q ^
0 c b
tj ^--CD

E |
o
O






Technology
Package
baseline
1
2
3
4
5
baseline
1
2
3
4
5
6
7
baseline
1
2
3
4
5
6
7
8
9
baseline
1
2
3
4
5
6
7
8

c

5 s i
Camshaft changes c „
(excluding those for £ ^ 5 ^
Engine Transmission System Voltage downsized engines) j if Q <
1.5L-4V DOHCI4 AT4spd 12V
1.5L-4VI4 AT4spd 12V LUB EFR
1.5L-4VI4CCP DCT 6 spd 12V LUB EFR
1.2LI3DWL+CCP + GDI dry DCT 6 spd 42 S-S LUB EFR I4tol3
0.7LI3 (small) Turbo DCP + GDI dry DCT 6 spd 42 S-S LUB EFR I4tol3
150kW/lithium ion (range of FTP 150 miles) N/A HEV
2.4L^tV DOHC 14 AT4spd 12V
2.4L4VI4 AT 4 spd 12V LUB EFR
2.0L 14 CCP + GDI AT6spd 12V LUB EFR I4tol4
2.4L-4VI4CCP + GDI DCT 6 spd 12V LUB EFR I4tol4
2.0L 14 DWL + CCP + GDI dry DCT 6 spd 42 S-S LUB EFR I4tol4
1. 5LI4 Turbo DCP+ GDI dry DCT 6 spd 42 S-S LUB EFR I4tol4
1.2LI4HEV(IMA)+GDI dry DCT 6 spd HEV LUB EFR I4tol4
1.2LI4HEVPIug-inlMA + GDI dry DCT 6 spd HEV LUB EFR I4tol4
2.4L^tV DOHC 14 AT4spd 12V
2.4L-4VI4 AT 4 spd 12V LUB EFR
2.2L 14 CCP + GDI AT 6 spd 12V LUB EFR I4tol4
2.2LI4DWL+CCP + GDI DCT 6 spd 12V LUB EFR I4tol4
2.2L 14 DWL + CCP + GDI dry DCT 6 spd 42 S-S LUB EFR I4tol4
1. 6LI4 Turbo DCP+ GDI dry DCT 6 spd 42 S-S LUB EFR I4tol4
1
^ tj in
^ % s D) $ V, ~
S»~ oigc |
: 
-------
Draft Regulatory Impact Analysis
                                                                Table 1-11 Continued

to
O
QJ
PEo
5 S~-
"QJ CD
•8
•o
E


a
QJ
ffl — .
It
O £°
QJ ^
•o
E


-g
> ,2
E 1_
T3 E t
<1) CO O
N =;
•35 S^
i-E
^ £
5.

-g

•~
"ro
E
•£,
£P
o C
-Q co
§5
>£
Q_
^
0)
.N
in
T3
E

baseline
1
2
3
4
5
6
7
8
baseline
1
2
3
4
5
6
7
8
9
baseline
1
2
3
4
5
6
7
8
9
baseline
1
2
3
4
5
6
7
8
9
10
11
12
13
3.3L-4V DOHC V6
3.3L-4VV6
3.0LV6GDI + CCP
3.0LV6w/DeacGDI + CCP
3.0LV6w/DeacGDI + CCP
2.2LI4 Turbo DCP+ GDI
3.0LV6HCCIGDI
2.2L Turbo DCP + GDI lean bum
2.5L 14 HEV (Power Split) + GDI
4.5L-4V DOHC V8
4.5L-4VV8
4.0LV6GDI + CCP
4.0LV6w/DeacGDI + CCP
4.0LV6w/DeacGDI + CCP
3.0LV6 Turbo DCP + GDI
3.0LV6 Turbo DCP + GDI
4.0LV6HCCIGDI
3.0LV6 Turbo Diesel
3.0L V6 w/ Deac GDI + CCP HEV (2-mode)
2.6L-4V DOHC 14 (15)
2.6L-4VI4
2.4LI4CCP+GDI
2.4LI4DWL + CCP + GDI
2.4LI4DWL + CCP + GDI
2.0LI4 Turbo DCP + GDI
1 .81 Turbo DCP + GDI lean bum
1.8L 14 Turbo HEV(IMA)+GDI
1.8L 14 Turbo HEV (Power Split) + GDI
1 .81 14 Turbo HEV Pluq-in IMA + GDI
3.7L-2VSOHCV6
3.7L 2V SOHC V6
3.2L 2V SOHC V6 GDI + CCP
3.2L 2V SOHC V6 w/Deac GDI + CCP
2.8L4VV6GDI +CCP
2.8L4VV6GDI +CCP+DWL
2.8L 4V V6 w/ Deac GO + CCP
2.8L 4V V6 w/ Deac GO + CCP
2.4L 14 Turbo DCP + GDI
2.8L4VV6HCCIGDI
2.4L 14 Turbo DCP +GDI lean bum
2.8L 14 Turbo Diesel
3.0L 4V V6 w/Deac GO +CCP HEV(IMA)
3. OL 4V V6 w/Deac GDI +CCP HEV (2-mode)
AT4spd
AT4spd
AT6spd
AT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
N/A
AT 4 spd
AT4spd
AT 6 spd
AT 6 spd
DCT6spd
AT 6 spd
DCT6spd
DCT6spd
DCT6spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT6spd
dry DCT 6 spd
dry DCT 6 spd
DCT 6 spd
dry DCT 6 spd
N/A
dry DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT6spd
DCT6spd
DCT6spd
DCT 6 spd
DCT 6 spd
N/A
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
HEV
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
HEV
HEV
HEV
12V
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
V6 SOHC to V6 DOHC LUB EFR
V6 SOHC to V6 DOHC LUB EFR
V6 SOHC to V6 DOHC LUB EFR
V6 SOHC to V6 DOHC LUB EFR
LUB EFR
V6 SOHC to V6 DOHC LUB EFR
LUB EFR
LUB EFR
V6 SOHC to V6 DOHC LUB EFR
V6 SOHC to V6 DOHC LUB EFR





V6DOHCtol4

V6DOHCtol4
V6DOHCtol4


V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC

V8 DOHC to V6 DOHC


14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14








V6 SOHC to 14

V6SOHCtol4




ASL
ASL
ASL



GDI-LB


ASL
ASL
ASL

ASL


Diesel-SCR


ASL
ASL



GDI-LB




ASL
ASL
ASL
ASL
ASL
ASL



GDI-LB
Diesel-SCR



TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
IACC 42V


TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 42V
IACC 42V
TORQ IACC 42V
IACC 42V
IACC 42V



TORQ IACC 12V
TORQ IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 42V




TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
IACC 42V





EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1


EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1


EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1
AERO1


EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR

$214
3% $1,022
5% $1,275
10% $2,103
10% $2,245
30% $3,754
30% $4,373
$6,005

$214
3% $883
5% $1,633
5% $1,719
5% $2,333
5% $2,419
30% $3,853
5% $3,456
$5,953

$214
3% $987
3% $1,255
10% $2,054
10% $2,369
30% $4,691
$4,628
$6,164
$14,226

$214
3% $1,044
3% $1,194
3% $1,398
5% $1,696
5% $1,665
10% $2,528
10% $2,555
30% $4,387
30% $4,890
5% $2,968
$4,710
$5,940
                                                                       1-14

-------
                                 Technology Packages, Cost and Effectiveness
Table 1-11 Continued


">;
0
.Q
'c _
5- o
Ig
;
o
.Q
i°
> p
i «)
s >

l>
— '



+ g
«?_
£|£

-------
Draft Regulatory Impact Analysis
                                                                Table 1-11 Continued


"£•
5
iff
1 B
t ^
«>
P
-1


^
It
H»
ll
„ «
H^
	
1
1
"O , — .
O Id
JH J-
l*t
g£i
Q_ CD
E >
QJ
.N
^
E


?~
.Q
^ S"
>^
% <°
s >
0)
e>
5

baseline
1
2
3
4
5
6
7
8
9
10
11
baseline
1
2
3
4
5
baseline
1
2
3
4
5
6
7
8
9
10
11
baseline
1
2
3
4
5
6
7
8
9
5.7L2VOHVV8
5.7L2VOHVV8
5.2L2VOHVV8GDI+CCP
5.2L 2V OHV V8 w/Deac GDI + CCP
4.6L4VV8GDI +CCP
4.6L 4V V8 w/ Deac GO + CCP
4.6L 4V V8 w/ Deac GO + CCP
3.5L 4V V6 Turbo DCP + GDI
4.6L4VV8HCCIGDI
3.5L V6 Turbo DCP + GDI lean burn
3. 5LV6 Turbo Diesel
4. 6L 4V V8 w/Deac GDI +CCP HEV(2-mode)
5.4L 3V SOHC V8
5.4L 3V SOHC - V8
4.6L 4V DOHC V8 GDI + CCP
4.6L 4V DOHC V8 w/ Deac GDI + CCP
4.6L 4V DOHC V8 w/ Deac GDI + CCP
3.5LV6 Turbo DCP + GO
3.2L-4V DOHC V6
3.2L-4VV6
2.8LV6GDI + CCP
2.8LV6GDI + CCP+DWL
2.8L V6 w/Deac GDI + CCP
2.8L V6 w/Deac GDI + CCP
2.4LI4 Turbo DCP + GDI
2.8LV6HCCIGDI
2.4L 14 Turbo DCP +GDI lean bum
2.8L 14 Turbo Diesel
3.0L V6 w/ Deac GDI + CCP HEV (IMA)
3.0L V6 w/ Deac GDI + CCP HEV (2-mode)
3.5L-4V DOHC V6
3.5L-4VV6
3.2LV6GDI + CCP
3.2L V6 w/Deac GDI + CCP
3.2L V6 w/Deac GDI + CCP
2.4LI4 Turbo DCP + GDI
3.2LV6HCCIGDI
2.4L 14 Turbo DCP +GDI lean bum
2.0L 14 Turbo HEV (IMA) + GDI
3.2L V6 w/ Deac GDI + CCP HEV (2-mode)
AT4spd
AT4spd
AT6spd
AT6spd
AT6spd
AT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
N/A
AT4sp
AT4spd
AT6spd
AT6spd
DCT6spd
DCT6spd
AT4spd
AT4spd
AT6spd
AT6spd
AT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
N/A
AT 4 spd
AT4spd
AT 6 spd
AT 6 spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
DCT6spd
N/A
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
42S-S
42S-S
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
HEV
HEV

LUB EFR
LUB EFR
LUB EFR
VSOHVtoVSDOHC LUB EFR
VSOHVtoVSDOHC LUB EFR
V8 SOHC to V8 DOHC LUB EFR
LUB EFR
VSOHVtoVSDOHC LUB EFR
LUB EFR
LUB EFR
VSOHVtoVSDOHC LUB EFR

LUB EFR
'SSOHCSVtoVSDOH LUB EFR
'SSOHCSVtoVSDOH LUB EFR
'SSOHCSVtoVSDOH LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

ASL
ASL
ASL
ASL
ASL

V8 OHV to V6 DOHC

V8 OHV to V6 DOHC GDI-LB
Diesel-SCR


ASL
ASL
ASL

V8 SOHC 3V to V6 DOHC

ASL
ASL
ASL
ASL

V6DOHCtol4

V6DOHCtol4 GDI-LB
Diesel-SCR



ASL
ASL
ASL

V6DOHCtol4

V6DOHCtol4 GDI-LB
V6DOHCtol4


TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
IACC 42V



TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
IACC 42V
IACC 42V

TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
IACC 42V




TORQ IACC 12V
TORQ IACC 12V
TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
IACC 42V




AERO1
AERO1
AERO1
AERO1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1


AERO1
AERO1
EPS AERO 1
EPS AERO 1


EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1


EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1

LRR
LRR
LRR
LRR
LRR
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB

LRR
LRR
LRR
LRR LDB
LRR LDB

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR


3%
3%
3%
5%
10%
10%
30%
30%
5%



3%
5%
10%
10%


3%
5%
5%
5%
5%
30%
30%
5%




3%
5%
10%
5%
30%
30%



$239
$1,051
$1,219
$1,559
$1,879
$2,913
$3,520
$5,647
$6,355
$3,844
$8,519

$239
$1,210
$1,530
$2,825
$3,295

$214
$1,044
$1,342
$1,311
$1,881
$2,023
$4,033
$4,494
$2,968
$4,356
$5,586

$214
$1,064
$1,345
$2,243
$2,057
$4,300
$4,919
$4,530
$6,095
                                                                       1-16

-------
                                                                                                              Technology Packages, Cost and Effectiveness
                                                                         Table 1-11 Continued

•>:
0
|p
> P
% «)
S >
0)
e>
3



"c"
>
is-
•g P
£s
«>
I1
_l



~
>
+ — ,
• — ' Ol
•s p
=5 C.
h= ro
«>
E>
(0
— 1

baseline
1
2
3
4
5
6
7
8
9
baseline
1
2
3
4
5
6
7
8
9
10
11
baseline
1
2
3
4
5
6
7
8
9
4.6L-4VDOHCV8
4.6L-4VV8
4.2LV6GDI + CCP
4.2L V6 w/ Deac GDI + CCP
4.2L V6 w/ Deac GDI + CCP
2.8LV6 Turbo DCP + GO
4.2LV6HCCIGDI
2.8L V6 Turbo DCP + GDI lean burn
3.0LV6 Turbo Diesel
4.2L V6 w/ Deac GDI + CCP HEV (2-mode)
4.0L-4V DOHC V6
4.0L-4VV6
3.6LV6GDI + CCP
3.6LV6GDI + CCP+DWL
3.6L V6w/ Deac GDI + CCP
3.6L V6w/ Deac GDI + CCP
2.5LI4 Turbo DCP + GDI
4.0LV6HCCIGDI
2.5L 14 Turbo DCP +GDI lean bum
2.8L 14 Turbo Diesel
3.6L V6 w/ Deac GDI + CCP HEV (IMA)
3.6L V6 w/ Deac GDI + CCP HEV (2-mode)
5.6L 4V DOHC V8
5.6L4VV8
4.6LV8GDI + CCP
4.6L V8 w/ Deac GDI + CCP
4.6L V8 w/ Deac GDI + CCP
3.5LV6 Turbo DCP + GO
4.6L V8 HCCI GDI
3.5L V6 Turbo DCP + GO lean burn
3. 5LV6 Turbo Diesel
4.6L V8 w/ Deac GDI + CCP HEV (2-mode)
AT4spd
AT4spd
ATSspd
ATSspd
DCT 6 spd
DCT6spd
DCT6spd
DCT6spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT6spd
DCT6spd
DCT6spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT6spd
DCT6spd
DCT6spd
DCT 6 spd
N/A
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR

LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR
LUB EFR


V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC

V8 DOHC to V6 DOHC






V6DOHCtol4

V6DOHCtol4








V8 DOHC to V6 DOHC

V8 DOHC to V6 DOHC



ASL TORQ IACC 12V
ASL TORQ IACC 12V
ASL TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
GDI-LB IACC 42V
Diesel-SCR


ASL TORQ IACC 12V
ASL TORQ IACC 12V
ASL TORQ IACC 12V
ASL TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
GDI-LB IACC 42V
Diesel-SCR



ASL TORQ IACC 12V
ASL TORQ IACC 12V
ASL TORQ IACC 12V
IACC 42V
IACC 42V
IACC 42V
GDI-LB IACC 42V
Diesel-SCR



EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1


EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1


AERO1
AERO1
EPS AERO 1
EPS AERO 1
EPS AERO 1
EPS AERO 1
AERO1
AERO1

LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR

LRR
LRR
LRR
LRR
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB

LRR
LRR
LRR
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB


3%
5%
10%
10%
30%
30%
5%



3%
5%
5%
10%
10%
30%
30%
5%




3%
5%
10%
10%
30%
30%
5%


$214
$926
$1,221
$2,153
$2,853
$4,415
$5,592
$3,037
$5,935

$239
$1,104
$1,488
$1,398
$2,381
$2,504
$4,613
$4,947
$3,871
$5,495
$8,010

$239
$1,051
$1,371
$2,515
$3,036
$5,139
$5,872
$3,844
$8,010
Notes to Table 1-11:

DOHC=dual overhead cam; SOHC=single overhead cam; OHV=overhead valve; AT=automatic transmission; DCT=dual clutch transmission; LUB=low friction lubes; EFR=engine friction
reduction; ASL=aggressive shift logic; TORQ=early torque converter lockup; IACC=improved accessories; EPS=electric power steering; AERO l=improved aerodynamics; LRR=low rolling
resistance tires.
                                                                                 1-17

-------
Draft Regulatory Impact Analysis
       Table 1-12 Package Costs & Effectiveness for 2012-2022+MY for 19 Vehicle Types (T1-T19), Packages Used as Inputs to the OMEGA Model, Costs in 2007 dollars





Vehicle

Q_
Jfl
-3 ^
CO ~


,_
8
0 f?
co ' —
v-^
o
Q


>
Q-
^
E >;„
22 o P
3 f s
N "

^


"ro ^
sa E
» s ?
u § t
O ^^ CD
s.? *
p
o





Technology
Package
baseline
1
2
3
4
5
baseline
1
2
3
4
5
6
7
baseline
1
2
3
4
5
6
7
8
9
baseline
1
2
3
4
5
6
7
8





Engine
1.5L-4VDOHCI4
1.5L-4VI4
1.5L-4VI4CCP
1.2LI3 DVVL+CCP + GDI
0.7LI3 (small) Turbo DCP+ GDI
150kW/lithium ion (range of FTP 150 miles)
2.4L-4V DOHC 14
2.4L4V 14
2.0LI4 CCP+GDI
2.4L -4V 14 CCP+GDI
2.0LI4 DVVL+CCP + GDI
1.5LI4 Turbo DCP +GDI
1.2LI4 HEV (I MA) + GDI
1.2LI4 HEVPIug-inlMA+GDI
2.4L-4V DOHC 14
2.4L-4VI4
2.2LI4 CCP+GDI
2.2LI4 DVVL+CCP + GDI
2.2LI4 DVVL+CCP + GDI
1. 6LI4 Turbo DCP + GDI
1 .6L Turbo DCP + GDI lean burn
1.4L 14 Turbo HEV (IMA) + GDI
1.8LI4 HEV (Power Split) + GDI
1.8LI4 HEV Plug-in Power Spl it + GDI
3.0L-4V DOHC V6
3.0L-4VV6
2.0L 14 Turbo DCP + GDI
2.0L 14 Turbo DCP + GDI
2.0L 14 Turbo DCP + GDI
2.0L Turbo DCP + GDI lean burn
2.4L 14 Turbo Diesel
1.5L 14 Turbo HEV (IMA) + GDI
2.8LV6w/DeacGDI +CCPHEV (2-mode)





Transm ission
AT 4 spd
AT 4 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
N/A
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A



Camshafi
System (excluding
Voltage downsizes
12V
12V
12V
42S-S
42S-S
HEV
12V
12V
12V
12V
42S-S
42S-S
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
HEV
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S

HEV
HEV



changes
those for
engines) 2012

$206
$801
$1,751
$2,343


$206
$911

$2,037




$206

$1,137
$2,108



$4,785


$231

$1,562
$2,392



$9,234





2013

$202
$779
$1,717
$2,291


$202
$886

$1,994




$202

$1,105
$2,063



$4,644


$227

$1,518
$2,341



$9,215





2014

$197
$757
$1,574
$2, 130


$197
$861

$1,843




$197

$1,074
$1,910



$4,506


$222

$1,475
$2, 168



$7,493





2015

$193
$736
$1,454
$1,994


$193
$837

$1,715




$193

$1,043
$1,780



$4,373


$218

$1,433
$2,021



$6,112





2016

$189
$716
$1,422
$1,946


$189
$814

$1,675




$189

$1,014
$1,739



$4,243


$214

$1,392
$1,974



$6,095





2017

$189
$716
$1,422
$1,946


$189
$814

$1,675




$189

$1,014
$1,739



$4,243


$214

$1,392
$1,974



$6,095





2018

$189
$716
$1,422
$1,946


$189
$814

$1,675




$189

$1,014
$1,739



$4,243


$214

$1,392
$1,974



$6,095





2019

$189
$716
$1,422
$1,946


$189
$814

$1,675




$189

$1,014
$1,739



$4,243


$214

$1,392
$1,974



$6,095





2020

$189
$716
$1,422
$1,946


$189
$814

$1,675




$189

$1,014
$1,739



$4,243


$214

$1,392
$1,974



$6,095





2021

$189
$716
$1,422
$1,946


$189
$814

$1,675




$189

$1,014
$1,739



$4,243


$214

$1,392
$1,974



$6,095





2022

$182
$685
$1,340
$1,820


$182
$784

$1,593




$182

$972
$1,656



$3,710


$207

$1,190
$1,725



$5,310
o
B
i!

s
o

7.6%
18.9%
33.2%
36.4%


7.6%
17.5%

35.4%




7.6%

23.2%
35.4%



35.9%


7.6%

23.4%
31.6%



36.5%
                                                                              1-18

-------
                                 Technology Packages, Cost and Effectiveness
Table 1-12 Continued
baseline
» 1
O 2
Q) Z
BE" 3
5 b 4
S-S 5
-o 6
E 7
8
baseline
3 1
8, 2
« - 3
_J CD
t K 4
0» 5
S3 6
i 7
E 8
9
baseline
% 1
£^ 2
E 15 3
»l§ 4
•?ls 5
^ E 6
E 1 7
a s
9
^ baseline
^ 1
t 2
» 3
<2 4
?j° 5
o O R
-Q CO b
§5 7
>g 8
1 9
S 10
•5 11
1 12
13
3.3L-4V DOHC V6
3.3L-4VV6
3.0LV6GDI+CCP
3.0L V6 w/ Deac GDI + CCP
3.0LV6w/DeacGDI+CCP
2.2L 14 Turbo DCP + GDI
3.0LV6HCCIGDI
2.2L Turbo DCP + GDI lean burn
2.5L 14 HEV (Power Split) + GDI
4.5L-4V DOHC V8
4.5L-4VV8
4.0LV6GDI+CCP
4.0L V6w/ Deac GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3.0LV6 Turbo DCP + GDI
3.0LV6 Turbo DCP + GDI
4.0LV6HCCIGDI
3.0LV6Turbo Diesel
3.0LV6w/DeacGDI + CCP HEV (2-mode)
2.6L-4VDOHCI4(I5)
2.6L-4VI4
2.4LI4 CCP + GDI
2.4LI4 DVVL+CCP + GDI
2.4LI4 DVVL+CCP + GDI
2.0L 14 Turbo DCP + GDI
1 .81 Turbo DCP + GDI lean burn
1.8L 14 Turbo HEV (IMA) + GDI
1.8L 14 Turbo HEV (Power Split) + GDI
1.8LI4 Turbo HEV Plug-in IMA + GDI
3.7L-2VSOHCV6
3.7L 2V SOHC V6
3.2L 2V SOHC V6 GDI + CCP
3.2L 2V SOHC V6 w/Deac GDI + CCP
2.8L 4V V6 GDI + CCP
2.8L 4V V6 GDI + CCP + DWL
2.8L 4V V6 w/ Deac GDI + CCP
2.8L 4V V6 w/ Deac GDI + CCP
2.4L 14 Turbo DCP + GDI
2.8L4V V6HCCIGDI
2.4L 14 Turbo DCP + GDI lean burn
2.8L 14 Turbo Diesel
3.0L 4V V6 w/ Deac GDI + CCP HEV (IMA)
3.0L 4V V6 w/ Deac GDI + CCP HEV (2-mode)
AT4spd
AT4spd
AT6spd
AT6spd
DCT 6 spd
DCT 5spd
DCT 5spd
DCT 5spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
DCT 6 spd
dry DCT 6 spd
N/A
dry DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 5spd
DCT 5spd
DCT 5spd
DCT 6 spd
DCT 6 spd
N/A
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
HEV
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
HEV
HEV
HEV
12V
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,144 $1,112 $1,081 $1,051 $1,022 $1,022 $1,022 $1,022 $1,022 $1,022 $927 17.9%

$2,537 $2,482 $2,305 $2,153 $2,103 $2,103 $2,103 $2,103 $2,103 $2,103 $1,942 34.2%



$6,772 $6,572 $6,377 $6,188 $6,005 $6,005 $6,005 $6,005 $6,005 $6,005 $5,129 37.5%

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$986 $959 $933 $907 $883 $883 $883 $883 $883 $883 $780 17.9%

$2,103 $2,061 $1,896 $1,757 $1,719 $1,719 $1,719 $1,719 $1,719 $1,719 $1,552 31.9%




$9,082 $9,068 $7,346 $5,966 $5,953 $5,953 $5,953 $5,953 $5,953 $5,953 $5,143 44.4%

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,104 $1,074 $1,044 $1,015 $987 $987 $987 $987 $987 $987 $890 17.4%
$1,406 $1,367 $1,328 $1,291 $1,255 $1,255 $1,255 $1,255 $1,255 $1,255 $1,139 21.4%
$2,481 $2,428 $2,252 $2,102 $2,054 $2,054 $2,054 $2,054 $2,054 $2,054 $1,896 34.7%
$2,837 $2,773 $2,587 $2,427 $2,369 $2,369 $2,369 $2,369 $2,369 $2,369 $2,172 36.3%


$6,951 $6,745 $6,545 $6,352 $6,164 $6,164 $6,164 $6,164 $6,164 $6,164 $5,335 39.6%


$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,168 $1,136 $1,104 $1,073 $1,044 $1,044 $1,044 $1,044 $1,044 $1,044 $948 17.8%
$1,337 $1,300 $1,263 $1,228 $1,194 $1,194 $1,194 $1,194 $1,194 $1,194 $1,093 19.6%
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC $3,017 $2,948 $2,756 $2,591 $2,528 $2,528 $2,528 $2,528 $2,528 $2,528 $2,354 32.4%

V6 SOHC to V6 DOHC


V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC $8,838 $8,807 $7,233 $5,968 $5,940 $5,940 $5,940 $5,940 $5,940 $5,940 $5,193 36.3%
       1-19

-------
Draft Regulatory Impact Analysis
                                                                Table 1-12 Continued
baseline
1
? 2
E 3
Isr 4
> H 5
|g 6
» 7
S 8
™ 9
10
11
baseline
1
? 2
E 3
is- 4
•> H 5
l» 6
»> 7
S 8
™ 9
10
11
baseline
1
2
e 3
« 4
i— 5
15 p 6
l± m 7
» > 8
S 9
-1 10
11
12
13
+ CD baseline
o > 1
£1? 2
» a) t 3
S> E> 4
ro ro H
J -1 5
4.0L-2VSOHCV6
4.0L 2V SOHC V6
3.6L 2V SOHC V6 GDI + CCP
3.6L 2V SOHC V6 w/Deac GDI + CCP
3.2L 4V V6 GDI + CCP
3.2L 4V V6 w/ Deac GDI + CCP
3.2L 4V V6 w/ Deac GDI + CCP
2.4L 14 Turbo DCP +GDI
3.2L4V V6HCCIGDI
2.4L 14 Turbo DCP +GDI lean burn
2.0L 14 Turbo HEV(IMA)+ GDI
3.2L 4V V6 w/ Deac GDI + CCP HEV (2-mode)
4.7L-2VSOHCV8
4.7L 2V SOHC V8
4.4L 2V SOHC V8 GDI + CCP
4.4L 2V SOHC V8 w/Deac GDI + CCP
4.2L 4V V6 GDI + CCP
4.2L 4V V6 w/ Deac GDI + CCP
4.2L 4V V6 w/ Deac GDI + CCP
2.8L 4V V6 Turbo DCP + GDI
4.2L4V V6HCCIGDI
2.8L 4V V6 Turbo DCP + GDI lean burn
3.0L V6 Turbo Diesel
4.2L 4V V6 w/ Deac GDI + CCP HEV (2-mode)
4.2L-2VSOHCV6
4.2L 2V SOHC V6
3.9L 2V SOHC V6 GDI + CCP
3.9L 2V SOHC V6 w/Deac GDI + CCP
3.6L 4V V6 GDI + CCP
3.6L 4V V6 GDI + CCP + DVVL
3.6L 4V V6 w/ Deac GDI + CCP
3.6L 4V V6 w/ Deac GDI + CCP
2.5L 14 Turbo DCP +GDI
4.0L4V V6HCCIGDI
2.5L 14 Turbo DCP + GDI lean burn
2.8L 14 Turbo Diesel
3.6L 4V V6 w/ Deac GDI + CCP HEV (IMA)
3.6L 4V V6 w/ Deac GDI + CCP HEV (2-mode)
3.8L2VOHVV6
3.8L-2V OHV V6
3.2L 4V DOHC V6 GDI + CCP
3.2L 4V DOHC V6 w/Deac GDI + CCP
3.2L4V DOHC V6 w/Deac GDI + CCP
2.5L 14 Turbo DCP + GDI
AT4spd
AT4spd
AT6spd
AT6spd
AT6spd
AT6spd
DCT 6 spd
DCT 5s pd
DCT 5s pd
DCT 6s pd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 5s pd
DCT 6s pd
DCT 5s pd
DCT 5 spd
N/A
AT 4 spd
AT 4 spd
AT 5 spd
AT 5 spd
AT 5 spd
AT 5 spd
AT 5 spd
DCT 5 spd
DCT 6s pd
DCT 6s pd
DCT 6s pd
DCT 6 spd
DCT 6 spd
N/A
AT4sp
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6s pd
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
HEV
HEV
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,191 $1,158 $1,126 $1,095 $1,064 $1,064 $1,064 $1,064 $1,064 $1,064 $969 17.4%
$1,360 $1,322 $1,285 $1,249 $1,214 $1,214 $1,214 $1,214 $1,214 $1,214 $1,113 19.4%
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
$3,125 $3,052 $2,857 $2,690 $2,623 $2,623 $2,623 $2,623 $2,623 $2,623 $2,409 32.3%
V6 SOHC to V6 DOHC


V6 SOHC to V6 DOHC $9,633 $9,602 $7,869 $6,477 $6,449 $6,449 $6,449 $6,449 $6,449 $6,449 $5,651 36.5%

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,215 $1,181 $1,148 $1,116 $1,085 $1,085 $1,085 $1,085 $1,085 $1,085 $990 17.4%
$1,384 $1,345 $1,307 $1,271 $1,235 $1,235 $1,235 $1,235 $1,235 $1,235 $1,134 19.4%


$2,999 $2,930 $2,739 $2,575 $2,512 $2,512 $2,512 $2,512 $2,512 $2,512 $2,342 34.3%




$9,458 $9,433 $7,704 $6,317 $6,294 $6,294 $6,294 $6,294 $6,294 $6,294 $5,486 36.5%

$256 $252 $247 $243 $239 $239 $239 $239 $239 $239 $231 7.6%
$1,233 $1,199 $1,167 $1,135 $1,104 $1,104 $1,104 $1,104 $1,104 $1,104 $984 18.3%
$1,424 $1,384 $1,346 $1,309 $1,273 $1,273 $1,273 $1,273 $1,273 $1,273 $1,146 19.9%
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC $3,256 $3,184 $2,977 $2,800 $2,735 $2,735 $2,735 $2,735 $2,735 $2,735 $2,529 34.9%

V6 SOHC to V6 DOHC


V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC

$256 $252 $247 $243 $239 $239 $239 $239 $239 $239 $231 7.6%
V6 OHV to V6 DOHC $1,482 $1,441 $1,401 $1,362 $1,324 $1,324 $1,324 $1,324 $1,324 $1,324 $1,196 18.9%
V6 OHV to V6 DOHC $1,672 $1,625 $1,580 $1,536 $1,493 $1,493 $1,493 $1,493 $1,493 $1,493 $1,358 20.1%
V6 OHV to V6 DOHC $3,381 $3,305 $3,095 $2,914 $2,845 $2,845 $2,845 $2,845 $2,845 $2,845 $2,635 34.9%
$3,653 $3,569 $3,351 $3,163 $3,087 $3,087 $3,087 $3,087 $3,087 $3,087 $2,820 35.1%
                                                                       1-20

-------
                                 Technology Packages, Cost and Effectiveness
Table 1-12 Continued


•£•
5
is-
"o I—
L= CO
»>
D)
to
_l


^r
8 b
H §
le
.5 >
+
=
£
CO
"C ^
E 2
'c ^ 1—
>^s
Q- cS
E >
N

E


§
E
3?
-^ jH
l»
^ >
Q)
D)
5

baseline
1
2
3
4
5
6
7
8
9
10
11
baseline
1
2
3
4
5
baseline
1
2
3
4
5
6
7
8
9
10
11
baseline
1
2
3
4
5
6
7
8
9
5.7L2VOHVV8
5.7L2VOHVV8
5.2L2VOHVV8 GDI+CCP
5.2L 2V OHV V8 w/Deac GDI + CCP
4.6L 4V V8 GDI + CCP
4.6L 4V V8 w/ Deac GDI + CCP
4.6L 4V V8 w/ Deac GDI + CCP
3.5L 4V V6 Turbo DCP + GDI
4.6L4V V8HCCIGDI
3.5L V6 Turbo DCP + GDI lean burn
3.5LV6Turbo Diesel
4.6L 4V V8 w/ Deac GDI + CCP HEV (2-mode)
5.4L 3V SOHC V8
5.4L 3V SOHC - V8
4.6L 4V DOHC V8 GDI + CCP
4.6L4V DOHC V8 w/Deac GDI+CCP
4.6L4V DOHC V8 w/Deac GDI+CCP
3.5LV6 Turbo DCP + GDI
3.2L-4V DOHC V6
3.2L-4VV6
2.8LV6 GDI+CCP
2.8LV6GDI+CCP+ DVVL
2.8L V6 w/ Deac GDI + CCP
2.8LV6 w/Deac GDI+CCP
2.4L 14 Turbo DCP + GDI
2.8LV6HCCIGDI
2.4L 14 Turbo DCP + GDI lean burn
2.8L 14 Turbo Diesel
3.0L V6 w/ Deac GDI + CCP HEV (IMA)
3. OLV6 w/Deac GDI + CCP HEV (2-mode)
3.5L-4V DOHC V6
3.5L-4VV6
3.2LV6 GDI+CCP
3.2LV6 w/Deac GDI+CCP
3.2L V6 w/ Deac GDI + CCP
2.4L 14 Turbo DCP + GDI
3.2LV6HCCIGDI
2.4L 14 Turbo DCP + GDI lean burn
2.0L 14 Turbo HEV (IMA) + GDI
3.2L V6 w/ Deac GDI + CCP HEV (2-mode)
AT4spd
AT4spd
AT6spd
AT6spd
AT6spd
AT6spd
DCT 6 spd
DCT 5spd
DCT 5spd
DCT 5spd
DCT 6 spd
N/A
AT4sp
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 5spd
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 5spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 5spd
DCT 5spd
DCT 5spd
DCT 6 spd
N/A
12V
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
42S-S
42S-S
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
HEV
HEV

$256 $252 $247 $243 $239 $239 $239 $239 $239 $239 $231 7.6%
$1,172 $1,141 $1,110 $1,080 $1,051 $1,051 $1,051 $1,051 $1,051 $1,051 $933 18.3%
$1,363 $1,325 $1,289 $1,253 $1,219 $1,219 $1,219 $1,219 $1,219 $1,219 $1,096 19.9%
V8 OHV to V8 DOHC
V8 OHV to V8 DOHC
V8 SOHC to V8 DOHC $3,457 $3,379 $3,167 $2,984 $2,913 $2,913 $2,913 $2,913 $2,913 $2,913 $2,705 34.9%
$4,142 $4,044 $3,812 $3,610 $3,520 $3,520 $3,520 $3,520 $3,520 $3,520 $3,223 35.1%
V8 OHV to V8 DOHC


V8 OHV to V8 DOHC

$256 $252 $247 $243 $239 $239 $239 $239 $239 $239 $239 7.6%
V8 SOHC 3V to V8 DOHC $1,352 $1,315 $1,279 $1,244 $1,210 $1,210 $1,210 $1,210 $1,210 $1,210 $1,161 18.9%
V8 SOHC 3V to V8 DOHC $1,714 $1,666 $1,619 $1,574 $1,530 $1,530 $1,530 $1,530 $1,530 $1,530 $1,456 21.2%
V8 SOHC 3V to V8 DOHC $3,357 $3,282 $3,073 $2,893 $2,825 $2,825 $2,825 $2,825 $2,825 $2,825 $2,687 34.9%
$3,889 $3,798 $3,573 $3,379 $3,295 $3,295 $3,295 $3,295 $3,295 $3,295 $3,085 35.1%

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,168 $1,136 $1,104 $1,073 $1,044 $1,044 $1,044 $1,044 $1,044 $1,044 $948 17.8%


$2,286 $2,239 $2,068 $1,924 $1,881 $1,881 $1,881 $1,881 $1,881 $1,881 $1,722 31.8%



$3,342 $3,244 $3,149 $3,057 $2,968 $2,968 $2,968 $2,968 $2,968 $2,968 $2,732 32.6%

$8,439 $8,420 $6,857 $5,603 $5,586 $5,586 $5,586 $5,586 $5,586 $5,586 $4,852 35.8%

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,191 $1,158 $1,126 $1,095 $1,064 $1,064 $1,064 $1,064 $1,064 $1,064 $969 17.4%


$2,485 $2,431 $2,255 $2,105 $2,057 $2,057 $2,057 $2,057 $2,057 $2,057 $1,806 31.6%



$9,234 $9,215 $7,493 $6,112 $6,095 $6,095 $6,095 $6,095 $6,095 $6,095 $5,310 36.5%
       1-21

-------
Draft Regulatory Impact Analysis
                                                                Table 1-12 Continued
baseline
S 1
1 2
Ip 3
~T H 4
a. ~~^ 5
?§ 6
B> 7
™ 8
9
baseline
1
e 2
§ 3
is- 4
"0 K 5
2^ 6
^ > y
E? Q
n 8
9
10
11
baseline
e 1
« 2
iffi- 3
o P 4
l± » 5
» > 6
if 7
-1 8
9
4.6L-4V DOHC V8
4.6L-4VV8
4.2LV6GDI+CCP
4.2L V6 w/ Deac GDI + CCP
4.2L V6w/ Deac GDI + CCP
2.8LV6 Turbo DCP + GDI
4.2LV6HCCIGDI
2.8L V6 Turbo DCP + GDI lean burn
3.0LV6 Turbo Diesel
4.2LV6w/DeacGDI + CCP HEV (2-mode)
4.0L-4V DOHC V6
4.0L-4VV6
3.6LV6GDI+CCP
3.6LV6GDI+CCP+ DVVL
3.6L V6 w/ Deac GDI + CCP
3.6L V6w/ Deac GDI + CCP
2.5L 14 Turbo DCP + GDI
4.0LV6HCCIGDI
2.5L 14 Turbo DCP + GDI lean burn
2.8L 14 Turbo Diesel
3.6L V6 w/ Deac GDI + CCP HEV (IMA)
3.6LV6w/DeacGDI + CCP HEV (2-mode)
5.6L4V DOHCV8
5.6L 4V V8
4.6LV8GDI+CCP
4.6L V8w/ Deac GDI + CCP
4.6L V8 w/ Deac GDI + CCP
3.5LV6 Turbo DCP + GDI
4.6LV8HCCIGDI
3.5L V6 Turbo DCP + GDI lean burn
3.5LV6 Turbo Diesel
4.6LV8w/DeacGDI + CCP HEV (2-mode)
AT 4 spd
AT 4 spd
ATSspd
ATSspd
DCT 6 spd
DCT 6spd
DCT 6spd
DCT 6spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6spd
DCT 6spd
DCT 6spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6spd
DCT 6spd
DCT 6spd
DCT 6 spd
N/A
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
12V
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV
HEV
12V
12V
12V
12V
42S-S
42S-S
42S-S
42S-S
12
HEV

$231 $227 $222 $218 $214 $214 $214 $214 $214 $214 $207 7.6%
$1,035 $1,006 $979 $952 $926 $926 $926 $926 $926 $926 $823 17.4%
$1,368 $1,329 $1,292 $1,256 $1,221 $1,221 $1,221 $1,221 $1,221 $1,221 $1,111 20.5%
$2,593 $2,537 $2,358 $2,205 $2,153 $2,153 $2,153 $2,153 $2,153 $2,153 $1,984 33.7%




$9,053 $9,040 $7,323 $5,948 $5,935 $5,935 $5,935 $5,935 $5,935 $5,935 $5,127 36.5%

$256 $252 $247 $243 $239 $239 $239 $239 $239 $239 $231 7.6%
$1,233 $1,199 $1,167 $1,135 $1,104 $1,104 $1,104 $1,104 $1,104 $1,104 $984 18.3%

$1,565 $1,521 $1,479 $1,437 $1,398 $1,398 $1,398 $1,398 $1,398 $1,398 $1,270 21.0%
$2,856 $2,796 $2,601 $2,436 $2,381 $2,381 $2,381 $2,381 $2,381 $2,381 $2,187 34.4%
$2,995 $2,931 $2,732 $2,562 $2,504 $2,504 $2,504 $2,504 $2,504 $2,504 $2,220 34.5%






$256 $252 $247 $243 $239 $239 $239 $239 $239 $239 $231 7.6%
$1,172 $1,141 $1,110 $1,080 $1,051 $1,051 $1,051 $1,051 $1,051 $1,051 $933 18.3%
$1,363 $1,325 $1,289 $1,253 $1,219 $1,219 $1,219 $1,219 $1,219 $1,219 $1,096 21.0%
$1,556 $1,515 $1,475 $1,435 $1,398 $1,398 $1,398 $1,398 $1,398 $1,398 $1,272 34.4%
$2,145 $2,086 $2,029 $1,973 $1,919 $1,919 $1,919 $1,919 $1,919 $1,919 $1,703 34.5%




                                                                       1-22

-------
                                          Technology Packages, Cost and Effectiveness
          Table 1-13 Package Costs Measured Relative to the Package Costs for the 2016MY
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022+
PACKAGE COSTS
RELATIVE TO 20 16
117%
115%
108%
102%
100%
100%
100%
100%
100%
100%
92%
       We do not show cost and effectiveness estimates for a number of the technology listed
in Table 1-12, as we determined that these packages were not cost effective relative to other
packages available for a specific vehicle type. The process used to make these determinations
is discussed below.

       As discussed in detail in Chapter 4 of this DRIA, the order of technology which will
be applied to any specific vehicle by the OMEGA model is set in the Technology input file.
Since the goal of adding technology is to move the manufacturer closer to compliance with
the GHG standard, the available technology packages should be placed in order of their total
GHG effectiveness.  Otherwise, the model is adding technology which moves the
manufacturer further from compliance. At the same time, the cost of each successive package
should be greater than that of the prior package. In this case, a greater degree of GHG
reduction is available at a lower cost. The package with the greater cost and lower overall
effectiveness  should therefore be removed from the list.

       Table  1-14 presents the  complete list of technology packages which were described
for vehicle type #6, which includes midsize and large cars equipped  with a V8 engine with
either SOHC or DOHC and 4 valves per head.  The only exception is that the package
including an HCCI engine has been removed as this technology is not expected to be
commercially available in time  for widespread introduction by 2016. The information listed
in the first six columns is taken from Table 1-12.  The values in the seventh column, which
are explained below, are used to remove packages which would not likely be applied by a
manufacturer and therefore, should not be included in the OMEGA modeling.
                                         1-23

-------
Draft Regulatory Impact Analysis
                Table 1-14 Evaluation of Technology Packages for Vehicle Type #6
Engine
4.5L DOHC 4-Valve V8
4.0L V6 GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3.0LV6TurboDCP + GDI
4.0L V6 w/ Deac GDI + CCP
3.0L V6 Turbo DCP + GDI
3. OLV6 Turbo Diesel
3.0L V6 w/ Deac GDI+CCP HEV
Transmission
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6spd
DCT 6 spd
2-mode
System
Voltage
12V
12V
42S-S
42S-S
42S-S
42S-S
12V
HEV
Weight
Reduction
0%
3%
5%
5%
5%
5%
5%
0%
Total CO2
Reduction
7.6%
17.9%
28.2%
28.5%
31.9%
32.1%
35.0%
44.4%
Total 20 16
Cost
$214
$883
$1633
$2333
$1719
$2419
$3456
$5953
$/delta CO2
%
$2,816
$6,497
$7,274
$284,727
$(17,617)
$2,363
$35,737
$26,586
Remove Turbo with AT 6 spd
4.5L DOHC 4-Valve V8
4.0L V6 GDI + CCP
4.0L V6 w/ Deac GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3.0LV6TurboDCP + GDI
3. OL V6 Turbo Diesel
3.0L V6 w/ Deac GDI + CCP HEV
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6spd
DCT 6 spd
2-mode
12V
12V
42S-S
42S-S
42S-S
12V
HEV
0%
3%
5%
5%
5%
5%
0%
7.6%
17.9%
28.2%
31.9%
32.1%
35.0%
44.4%
$214
$883
$1633
$1719
$2419
$3456
$5953
$2,816
$6,497
$7,274
$2,305
$452,082
$35,737
$26,586
Remove Turbo with DCT 6 spd
4.5L DOHC 4-Valve V8
4.0L V6 GDI + CCP
4.0L V6 w/ Deac GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3. OLV6 Turbo Diesel
3.0L V6 w/ Deac GDI + CCP HEV
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
2-mode
12V
12V
42S-S
42S-S
12V
HEV
0%
3%
5%
5%
5%
0%
7.6%
17.9%
28.2%
31.9%
35.0%
44.4%
$214
$883
$1633
$1719
$3456
$5953
$2,816
$6,497
$7,274
$2,305
$56,828
$26,586
Remove Diesel
4.5L DOHC 4-Valve V8
4.0L V6 GDI + CCP
4.0L V6 w/ Deac GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3.0L V6 w/ Deac GDI + CCP HEV
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
2-mode
12V
12V
42S-S
42S-S
HEV
0%
3%
5%
5%
0%
7.6%
17.9%
28.2%
31.9%
44.4%
$214
$883
$1633
$1719
$5953
$2,816
$6,497
$7,274
$2,305
$34,012
Remove two AT 6 spd steps
4.5L DOHC 4-Valve V8
4.0L V6 w/ Deac GDI + CCP
3.0L V6 w/ Deac GDI + CCP HEV
AT 4 spd
DCT 6 spd
2-mode
12V
42S-S
HEV
0%
5%
0%
7.6%
31.9%
44.4%
$214
$1719
$5953
$2,816
$6,184
$34,012
       The seventh, or last column of Table 1-14 is a measure of the incremental cost
effectiveness of each package relative to the previous package.  Specifically, it is the ratio of
the incremental cost of the current package over the previous package to the incremental
effectiveness of the current package over the previous package.  In both cases (cost and
effectiveness), the increment is the arithmetic difference. As discussed above, OMEGA uses
a different measure of incremental effectiveness in its calculation of CO2 emissions. Here,
however, the arithmetic difference in the effectiveness of two technology packages provides
the best comparison across packages, since the base CO2 emissions inherent in the total
effectiveness estimates is the same; that of the base vehicle.  Therefore, a 10% difference
between two packages with 7% and 17% effectiveness, respectively, represents the same CO2
emission reduction as a 10% difference between two packages with 27% and 37%
                                            1-24

-------
                                          Technology Packages, Cost and Effectiveness
effectiveness, respectively.  Generally, a low ratio of incremental cost to incremental
effectiveness is better than a high ratio. Ideally, the technology packages included in the
model would progress from lower ratios to higher ratios.

       The topmost section of Table 1-14 shows all of the packages described earlier except
for the HCCI engine package. The order of the packages has been rearranged slightly from
that in Table 1-12 in order to place the packages in order of increasing total effectiveness.  As
can be seen, there are two very large anomalies in the ratios of incremental cost to incremental
effectiveness.  The ratio for the turbocharged engine with a 6 speed automatic transmission is
very high, while that for the engine with cylinder deactivation with a dual clutch transmission
is negative. The cause of this is that the cost of the latter package is lower than the former. If
the latter package can achieve a 31.9% reduction in €62 emissions at a cost of $1,719, then
there is no point in considering a package which only achieves a 28.5% reduction in CCh
emissions for a cost of $2,333. Therefore, we removed the package for the turbocharged
engine with a 6 speed automatic transmission and repeated the calculations. (In general, the
package just prior to one with a negative ratio of incremental cost to incremental effectiveness
should be removed.)  The revised set of technology packages is shown in the second section
of Table  1-14.

       The second set of packages now shows one obvious anomaly. The ratio of
incremental cost to incremental effectiveness for the turbocharged engine with a dual clutch
transmission is more than a factor of 10 higher than any of the others.  This occurs because
this package only reduces CCh emissions by 0.2% over the previous package for an
incremental cost of $700. A manufacturer would be better off applying the next package (the
diesel) to a portion of its vehicles than to  apply the turbocharged engine with a dual clutch
transmission package to a higher percentage of its vehicles. Therefore, we removed the
package for the turbocharged engine with the dual clutch transmission and again repeated the
calculations. The revised set of technology packages is shown in the third section of Table
1-14.

       The greatest anomaly in the third set of ratios is that for the diesel package.  It is more
than twice the value of the ratio for the 2-mode hybrid.  If we believed that manufacturers
would prefer to implement diesel technology over strong hybridization for some reason, we
could have left both packages in the modeling. However, absent such a reason, we removed
the diesel engine package from vehicle type #6.  The revised set of technology packages is
shown in the fourth section of Table 1-14.

       The greatest anomaly in the fourth section of the table is that the ratio of incremental
cost to incremental effectiveness for the engine with cylinder deactivation and a dual clutch
transmission is much lower than those for the two prior packages. In order to assess the
benefit of removing the two prior packages, we do so in the fifth and last section of Table
1-14. As can be seen, the ratio of incremental cost to incremental effectiveness for the engine
with cylinder deactivation and a dual clutch transmission increases to over $6,184.  This is
only marginally lower than the ratios for the two packages which have been excluded.
Retaining the two packages  provides a more gradual application of technology. This provides
the model the choice of applying technology which is currently widespread in the fleet (6-
speed automatic transmissions and 12 volt electrical systems) to a greater percentage of sales

                                        1-25

-------
Draft Regulatory Impact Analysis
before applying more extensive technology (dual clutch transmissions and start stop
technology). Therefore, we did not exclude the two technologies based on a strict use of the
ratio of incremental cost to incremental effectiveness.
1.4  EPA's  Lumped Parameter Approach  for Determining  Effectiveness
     Synergies

       EPA engineers reviewed existing tools that could be used to develop estimates of the
technology synergies, including the NEMS model.:  However, the synergies in the NEMS
model depend heavily upon an assumed technology application flow path; those technologies
that the model would apply first would be expected to have fewer synergies than those applied
later on. For this reason, and because this report includes many new technologies not
available in NEMS, it was necessary for EPA to develop its own set of estimates. EPA used a
well-documented engineering approach known as a lumped-parameter technique to determine
values for synergies. At the same time, however, EPA recognized the availability of more
robust methods for determining the synergistic impacts of multiple technologies on vehicle
CO2 emissions than the lumped-parameter approach, particularly with regard to applying
synergy effects differentiated across different vehicle classes, and therefore augmented this
approach with the detailed vehicle simulation modeling described in Section 1.4.7.

       The basis for EPA's lumped parameter analysis is a first-principles energy balance
that estimates the manner in which the chemical energy of the fuel is converted into various
forms of thermal and mechanical energy on the vehicle.  The analysis accounts for the
dissipation of energy into the different categories of energy losses, including each of the
following:

       •  Second law losses (thermodynamic losses inherent in the combustion of fuel),
       •  Heat lost from the combustion process to  the exhaust and coolant,
       •  Pumping losses, i.e., work performed by  the engine during the intake and exhaust
          strokes,
       •  Friction losses in the engine,
       •  Transmission losses, associated with friction and other parasitic losses,
       •  Accessory  losses, related directly to  the parasitics  associated with the engine
          accessories and indirectly to the fuel efficiency losses related to engine warmup,
       •  Vehicle road load (tire and aerodynamic)  losses;

with the remaining energy available to propel the vehicle.  It is assumed that the baseline
vehicle has a fixed percentage of fuel lost to each category.

       Each technology is categorized into the major types of engine losses it reduces, so  that
interactions between multiple technologies applied to the vehicle may be determined. When a
technology is applied, its effects are estimated by modifying the appropriate loss categories by
a given percentage. Then, each subsequent technology that reduces the losses in an already
improved category has less of a potential impact than it would if applied on its own. Table
                                            1-26

-------
                                                 Technology Packages, Cost and Effectiveness
1-15 below is an example spreadsheet used by EPA to estimate the synergistic impacts of a
technology package for a standard-size car.
                         Table 1-15 Sample Lumped Parameter Spreadsheet

                                   EPA Staff Deliberative Materials-Do Not Quote or Cite
                            Vehicle Energy Effects Estimator
   Vehicle type: Standard Car
   Family
      Description: Technology picklist
      Package: Z





Baseline % of fuel
Reduction
% of original fuel


Vehicle
Mass
Inertia
Load
13.0%
0%
13.0%
In
Brake Energy
Road Loads
Drag Tires
Aero Rolling
Load Load
4.0% 4.0%
16% 8%
3.4% 3.7%
icated Ener

Parasitics

Access
Losses
1.8%
64%
0.8%
?y

Gearbox,
T.C.
Losses
4.2%
33%
3.3%

Engine Friction


Friction Pumping
Losses Losses
6. 6% 4.4%
16% 75%
5.6% 1.1%
Heat
Lost To
Exhaust &
Coolant
IndEff
Losses
32.0%

31.8%




Second
Law
30.0%

30%
                                                                           Check
                                                                           100.0%
        Current Results
    72.9%  Fuel Consumption
    27.1%  FC Reduction
    37.2%  FE Improvement
     N/A  Diesel FC Reduction
Original friction/brake ratio
Based on PMEP/IMEP »»
(GM study)
Technology
Aero Drag Reduction
Rolling Resistance Reduction
LowFric Lubes
EF Reduction
ICP
DCP
CCP
Deac
DWL
CWL
Cam! ess
GDI
Turbo/Dnsrze
5-spd
CVT
ASL
Agg TC Lockup
6-spd auto
AMT
42V S-S
12V ace + Imp alt
EPS
42V ace + imp alt
HCCI dual-mode
GDI (lean)
Diesel - LNT
Diesel - SCR
Opt. E25
Independent
FC Estimate
3.0%
1.5%
0.5%
2.0%
2.0%
3.0% tota
3.0% tota
6.0%
4.0%
5.0%
10.0%
1.5%
6.0%
2.5%
6.0%
1.5%
0.5%
5.5%
6.5%
7.5%
1.5%
1.5%
3.0%
11.0%
10.5%
30.0% ove
30.0% ove
8.5%
Loss Category
Aero
Rolling
Friction
Friction
Pumping
WT Pumping
WT Pumping
Pumping, friction
Pumping
Pumping
Pumping
IndEff
Pumping
Trans, pump ing
Trans, pumping
Pumping
Trans
Trans, pumping
Trans
F,P,A
Access
Access
Access
nd. Eff, pumping
nd. Eff, pumping
r gas nd Eff, pumping
r gas nd Eff, pumping
nd. Eff, pumping
Implementation into estimator
16% aero (cars), 10.5% aero (trucks)
8% rolling
2% friction
8.5% friction
12%pumping, 38.2% IE, -2%fric
18.5% pumping, 38.2% IE, -2%fric
18.5% pumping, 38.2% IE, -2%fric
39% pumping
30%pumping, -3%friction
37%pumping, -3%friction
76% pumping, -5% friction
38. 6% IndEff
39% pumping
22% pumping, -5% trans
46% pumping, -5% trans
9.5% pumping
2. 5% trans
42% pumping, -5% trans
35% trans (increment)
13% friction, 19% pumping, 38% access
18% access
18% access
3 6% access
41%IE, 25% pump ng
40% IE, 38% pump ng
48% IE, 85% pump ng, -13% friction
46% IE, 80% pump ng, -13% friction
39% IE, 40% pump ng
User Picklist
Include? (0/1)
1
1
1
1
0
0
1
0
1
0
0
0
0
0
0
1
1
1
1
1
0
1
1
0
0
0
0
0
                                               1-27

-------
Draft Regulatory Impact Analysis
       Table 1-16 below lists the technologies considered in this example, their
corresponding individual technology effectiveness values, and a comparison of the gross
combined package CCh reduction (i.e. disregarding synergies) to the lumped parameter
results.  The difference is the implied synergistic effects of these technologies combined on a
package.

         Table 1-16 Comparison of Lumped Parameter Analysis with Standard Car Package
TECHNOLOGY
Aero Drag
Rolling Resistance Reduction
Low Friction Lubricants
Engine Friction Reduction
WT - Coupled Cam Phasing
WT - Discrete Variable Lift
Aggressive Shift Logic
Early Torque Converter Lock-up
6-speed Automatic Transmission
6-speed Dual Clutch Transmission
Stop-start with 42 volt system
Electric Power Steering
42V ace + improved alternator
Gross combined effectiveness
Lumped Parameter Estimate
Estimated synergistic effects
INDIVIDUAL CO2
REDUCTION
3%
1.5%
0.5%
2.0%
3.0%
4.0%
1.5%
0.5%
5.5%
6.5%
7.5%
1.5%
3.0%
33.6%
27.1%
-6.5%
CUMULATIVE CO2
REDUCTION
3%
4.5%
4.9%
6.8%
9.6%
13.2%
14.5%
15.0%
19.6%
24.9%
30.5%
31.5%
33.6%

       The synergy estimates obtained using the lumped parameter technique were
subsequently compared to the results from the vehicle simulation work. EPA will continue to
use the lumped parameter approach as an analytical tool, and (using the output data from the
vehicle simulation as a basis) may adjust the synergies as necessary in the future.
1.4.1
Ricardo's Vehicle Simulation
       Vehicle simulation modeling was performed by Ricardo, Inc. The simulation work
addressed gaps in existing synergy modeling tools, and served to both supplement and update
the earlier vehicle simulation work published by NESCCAF.  Using a physics-based, second-
by-second model of each individual technology applied to various baseline vehicles, the
Ricardo model was able to estimate the effectiveness of the technologies acting either
individually or in combination. This information could then be  used to estimate the synergies
of these technology combinations, and also to differentiate the synergies across different
vehicle classes.
                                            1-28

-------
                                          Technology Packages, Cost and Effectiveness
       In total, Ricardo modeled five baseline vehicles and twenty-six distinct technology
combinations, covering the full range of gasoline and diesel powertrain technologies used in
the Volpe model, with the exception of the powersplit, plug-in and two-mode hybrid vehicle
technologies.  The five generalized vehicle classes modeled were a standard car, a full-size
car, a small multi-purpose vehicle  (MPV), a large MPV and a large truck. The complete list
of vehicles and technology packages is given below in this section, along with a detailed
explanation of the selection criteria.

       Each technology package was modeled under a constraint of "equivalent
performance" to  the baseline vehicle. To quantify the performance, a reasonably
comprehensive, objective set of vehicle performance criteria were used as a basis to compare
with the baseline vehicle, characterizing the  launch acceleration, passing performance and
grade capability that a vehicle buyer might expect when considering a technology package.
The main metrics used to compare vehicle performance are listed below in Table 1-17.
            Table 1-17 Performance Metrics Used as Basis for "Equivalent Performance"
CHARACTERISTIC
Overall Performance
Launch Acceleration
Passing Performance
Grade Capability
PERFORMANCE METRIC
Time to accelerate from 0-60 MPH
Time to accelerate from 0-30 MPH
Vehicle speed and distance after a 3 -second acceleration
from rest
Time to accelerate from 30 to 50 MPH
Time to accelerate from 50 to 70 MPH
Maximum % grade at 70 MPH
(standard car, large car, small MPV and large MPV)
Maximum % grade at 60 MPH at GCVWR (large truck)
Notes: All accelerations are assumed at WOT (wide open throttle) condition. GCVWR = Gross Combined
Vehicle Weight Rating

       A summary of the vehicle simulation results is given below in Section 1.4.7, including
the CO2 emissions reduction effectiveness for each technology package. The full Ricardo
vehicle simulation results, including the acceleration performance data, may be found in
Ricardo's final report posted publicly at EPA's website.2

1.4.2  Description of Ricardo's Report

       In this section, the structure, methodology and results from the Ricardo vehicle
simulation report are summarized. EPA worked closely with Ricardo to develop baseline
models of five generalized vehicle classes that could be validated against EPA certification
data, and then used as a platform upon which to add various technology packages.  The
vehicle simulation modeling results generated by Ricardo consist of the following:

    •   Baseline vehicle characterization, to determine the baseline fuel consumption and CO2
       emissions over the EPA combined cycle federal test procedure (FTP) for five baseline
       vehicles, for validation with EPA certification data.
    •   Simulation of the vehicle technology combinations (applied to the baseline vehicles)

                                         1-29

-------
Draft Regulatory Impact Analysis
    •   Incremental technology  effectiveness  estimates, to examine the effect  of adding
       technologies one-by-one. These could then be used more directly to validate synergies
       estimated using the lumped parameter method.
       This section describes the selection process for each of the baseline vehicles and the
technology packages, and summarizes the results of the vehicle simulation.

1.4.3 Determination of representative vehicle classes

       In an effort to establish a reasonable scope for the vehicle simulation work and to
update the earlier simulation done by NESCCAF, EPA chose five representative vehicle
classes as the basis for evaluating technology benefits and synergies, representing the vehicle
attributes of the projected highest-volume light-duty car and truck sales segments. These five
classes covered a broad range of powertrain and vehicle characteristics, over which the
effectiveness and synergies of each of the technologies could be evaluated.  The main
distinguishing attributes of the five vehicle classes considered by EPA and Ricardo are given
below in Table 1-18.
         Table 1-18 Attributes of the Five Generalized Vehicle Classes Considered by Ricardo
VEHICLE
CLASS
EPA Vehicle
Types Included
Curb Weight
Range
Engine Type
Drivetrain
Body Type
Towing
Capability
Example vehicles
STANDARD
CAR
Compact,
Midsize
2800-3600 Ibs
14
FWD
Unibody
None
Toyota Camry,
Chevy Malibu,
Honda Accord
LARGE
CAR
Large CAR
>3600 Ibs
V6
RWD/AWD
Unibody
None
Chrysler 300,
Ford 500 /
Taurus
SMALL
MPV
Small SUV,
Small
Pickup
3600-4200
Ibs
14
FWD
Unibody
Partial
Saturn Vue,
Ford
Escape,
Honda CR-
V
LARGE
MPV
Minivans,
Mid-SUV's
4200-4800 Ibs
V6
FWD/AWD
Unibody
Partial
Dodge Grand
Caravan,
CMC Acadia,
Ford Flex
LARGE
TRUCKS
Large SUV's,
Large Pickups
>4800 Ibs
V8
4WD
Ladder Frame
Full
Ford F- 150,
Chevy
Silverado
1500, Dodge
Ram
EPA then selected representative vehicle models for each of these classes, based on three
main criteria:

    •   The vehicle should possess major attributes  and technology characteristics that are
       near the  average  of its class, including engine type and displacement,  transmission
       type, body type, weight rating, footprint size and fuel economy rating.
                                             1-30

-------
                                          Technology Packages, Cost and Effectiveness
    •   It should be  among the  sales volume leaders  in its class,  or where there  is not a
       clearly-established volume leader, the model should share attributes consistent with
       major sellers.

    •   The  vehicle  should have  undergone a recent update or  redesign, such  that the
       technology in the baseline model could be considered representative of vehicles sold
       at the beginning of the proposed regulatory timeframe.

       Consideration was also given to include the sales-leading vehicle manufacturers
among the baseline models.  Hence, the U. S. domestic manufacturers account for four of the
five models  (Chrysler 300, GM/Saturn Vue, Chrysler/Dodge Caravan, and the Ford F-150),
while import manufacturers are represented in their strongest sales segment, the standard car
class, by the Toyota Camry.

1.4.4 Description of Baseline Vehicle Models

       The baseline vehicles selected to represent their respective vehicle classes are
described below in Table 1-19, listed with the critical attributes that EPA used as  selection
criteria. While each attribute for these baseline vehicles does not match the precise average
for its class, each of these baselines is an actual vehicle platform that allows validation of the
simulation data with "real world" certification data.
                                         1-31

-------
Draft Regulatory Impact Analysis
                         Table 1-19 Description of Baseline Vehicles
VEHICLE CLASS
Baseline Vehicle
CC>2 Emissions*
(g/mi)
Vehicle Attributes
Performance
Characteristics
Base Engine
Displacement
(L)
Rate Power
(HP)
Torque (ft-lbs)
Valvetrain Type
Valves/Cylinder
Drivetrain
Transmission
# of Forward
Speeds
Curb Weight
Obs)
ETW Qbs)
GVWR Qbs)
GCWR Obs)
Front Track
Width (in.)
Wheelbase (in.)
Displacement /
Weight Ratio
(L/ton)
Power /
Weight Ratio
(HP/ton)
STANDARD
CAR
Toyota
Camry
327
DOHC 14
2.4
154
160
VVT (DCP)
4
FWD
Auto
5
3108
3500
—
—
62
109.3
1.54
99.1
LARGE
CAR
Chiysler 300
409
SOHC V8
3.5
250
250
Fixed
4
RWD
Auto
5
3721
4000
—
—
63
120
1.88
134.4
SMALL
MPV
Saturn
VUE
415
DOHC 14
2.4
169
161
VVT (DCP)
4
FWD
Auto
4
3825
4000
4300
—
61.4
106.6
1.25
88.4
LARGE
MPV
Dodge Grand
Caravan
435
OHVV6
3.8
205
240
Fixed
2
FWD
Auto
4
4279
4500
5700
—
63
119.3
1.78
95.8
LARGE
TRUCKS
Ford F- 150
575
SOHC V8
5.4
300
365
VVT (CCP)
3
4WD
Auto
4
5004
6000
6800
14000
67
144.5
2.16
119.9
       *-Estimated CCh equivalent, taken from EPA adjusted combined fuel economy
ratings.
1.4.5 Technologies Considered by EPA and Ricardo in the Vehicle Simulation

       A number of advanced gasoline and diesel technologies were considered in the
Ricardo study, comprising the majority of the technologies used in the Volpe model, with the
exception of the hybrid electric vehicle technologies.  In developing a comprehensive list of
technologies to be modeled, EPA surveyed numerous powertrain and vehicle technologies
and technology trends, in order to assess their potential feasibility in the next one to ten years.
                                           1-32

-------
                                           Technology Packages, Cost and Effectiveness
The list of technologies considered therefore includes those that are available today (e.g.,
variable valve timing, six-speed automatic transmissions) as well as some that may not be
ready for five to ten years (e.g., camless valve actuation and HCCI engines). Table 1-20
below lists the technologies that Ricardo included in the vehicle simulation models.

               Table 1-20 Technologies Included in the Ricardo Vehicle Simulation
ENGINE TECHNOLOGIES
Abbreviation
DOHC
SOHC
OHV
CCP
DCP
DVVL
CVVL
Deac
CVA
Turbo
GDI
Diesel
HCCI
LUB
EFR
Description
Dual Overhead Camshafts
Single Overhead Camshaft
Overhead Valve (pushrod)
Couple Cam Phasing
Dual (independent) Cam Phasing
Discrete (two-step) Variable Valve Lift
Continuous Variable Valve Lift
Cylinder Deactivation
Camless Valve Actuation (full)
Turbocharging and engine downsizing
Gasoline Direct Injection
Diesel with advanced aftertreatment
Homogenous Charge Compression Ignition (gasoline)
Low-friction engine lubricants
Engine Friction Reduction
TRANSMISSION TECHNOLOGIES
Abbreviation
L4
L5
L6
DCT6
CVT
ASL
TORQ
Description
Lock-up 4-speed automatic transmission
Lock-up 5-speed automatic transmission
Lock-up 6-speed automatic transmission
6-speed Dual Clutch Transmission
Continuously Variable Transmission
Aggressive Shift Logic
Early Torque Converter Lock-up
ACCESSORY TECHNOLOGIES
Abbreviation
ISG (42V)
EPS
EACC
HEA
Description
42V Integrated Starter-Generator
Electric Power Steering
Electric Accessories (water pump, oil pump, fans)
High-Efficiency Alternator
VEHICLE TECHNOLOGIES
Abbreviation
AERO
ROLL
Description
Aerodynamic drag reduction (10-20%)
Tire Rolling Resistance reduction (10%)
                                          1-33

-------
Draft Regulatory Impact Analysis
1.4.6 Choice of Technology Packages

       EPA chose a number of technology packages representing a range of options that
manufacturers might pursue. In determining these technology combinations, EPA considered
available cost and effectiveness numbers from the literature, and applied engineering
judgment to match technologies that were  compatible with each other and with each vehicle
platform. Also, where appropriate, the same technologies were applied to multiple vehicle
classes, to determine where specific vehicle attributes might affect their benefits and
synergies. Table 1-21 below describes in detail the technology content in each technology
package simulated by Ricardo.
           Table 1-21 Description of the Vehicle Technology Packages Modeled by Ricardo
VEHICLE
CLASS
Standard
Car
>
~S
CO
s*
03
u
01
ES9
33
tu
>
a
E
03
hJ

-------
                                            Technology Packages, Cost and Effectiveness


11
12
17
XI
X2
Turbo
4.8L V8, Diesel
5.4LV8, GDI
5.4L V8, GDI
5.4L V8, GDI
5.4LV8, GDI,
HCCI

DOHC
CCP, Deac
DCP, DVVL
CVA
DCP, CVVL

DCT6
L6
L6
DCT6
DCT6

EPS, EACC, HEA
ISO (42V), EPA,
EACC
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
       Other: 20% Aerodynamic drag reduction, 10% tire rolling resistance reduction assumed for all
vehicles, except Large Trucks. 10% Aerodynamic drag reduction assumed for Large Truck. Low-Friction
lubricants and moderate engine friction reductions are assumed for all vehicles. Aggressive shift logic and early
torque converter lockup strategies are assumed for all vehicles, where applicable.
1.4.7 Simulation Results

       The CO2 emissions results from the vehicle simulation are summarized below in Table
1-22 (for cars) and Table 1-23 (for light-duty trucks). The CO2 estimates are given for the
combined city and highway test cycles, according to the EPA Federal Test Procedure (FTP),
with the technology package results compared with the baseline vehicle as shown.

       It is important to reiterate that each of the technology package results were obtained
with performance determined to be equivalent to the baseline vehicle.  No attempt was made
to project trends in performance during the proposed regulatory period, nor was the
performance downgraded to give improved fuel efficiency. A full comparison of vehicle
acceleration performance is given in the Ricardo final report.
            Table 1-22 CO2 Emissions Estimates Obtained from Vehicle Simulation (Cars)
VEHICLE
Standard Car
0»
N
® $
==u
tu
TECHNONOLGY
PACKAGE
Baseline
Z
1
2
Baseline
4
5
Yl
MAJOR
FEATURES*
2.4L 14, DCP, L5
CCP, DVVL,
DCT, ISO
GDI, DCP,
DVVL, CVT
GDI, DCP, L6,
ISO
3.5L V6, L5
2.2L 14, GDI,
Turbo, DCP, L6
2.8L 14 Diesel,
DCT
GDI, CVA, DCT
C02
CITY
g/mi
338
250
294
277
420
346
315
278
C02
HWY
g/mi
217
170
198
180
279
236
221
199
C02
COMB
g/mi
284
214
251
233
356
296
273
242
C02
REDUCTION
%
—
24.7%
11.5%
17.8%
—
16.9%
23.5%
32.0%
                                           1-35

-------
Draft Regulatory Impact Analysis

Y2
6a
16
GDI, HCCI, DCT
GDI, DCP,
CVVL, DCT
GDI, CCP, Deac,
L6, ISO
290
331
301
197
235
205
248
288
257
30.4%
19.2%
27.7%
       *-Please refer to Table 1-20 for a full description of the vehicle technologies
     Table 1-23 CO2 Emissions Estimates Obtained from Vehicle Simulation (Light-Duty Trucks)
VEHICLE
PH
s
03
>

-------
                                          Technology Packages, Cost and Effectiveness
    1)  EPA's lumped-parameter package estimates are comparable with those obtained from
       the detailed Ricardo simulations. This is illustrated in Figure 1-2 below.
    2)  EPA  is  confident  in  the  plausibility  of the individual technology effectiveness
       estimates in, based on the  sources from which that information was assimilated, as
       detailed in Section 2 of this report.
    3)  Additionally, EPA expresses confidence in the overall Ricardo package results due to
       the robust methodology used in building the models and generating the results.





_


a a s
o o c
5 5 5
W W c/
Z 2

r
r.
r
r
r


>>>>>>>aaaaa
PHPHPHPHPHPHPnOOOOO
esssssslllll
WWWWWWW^^^JJ
Z 1 2 15 5a 15b 5 4 5 6a 16 Yl
Package

r

n
ri



3£££|||||||
»???HHHHHHH
a ton bii bii
J 3 3 3
Y2 4 6b 16 9 10 11 12 17 XI X2
   Figure 1-2 Comparison of Ricardo package results to equivalent lumped parameter package results

       Based on this, EPA concludes that the synergies derived from the lumped parameter
approach are generally plausible (with a few packages that garner additional investigation).
EPA will continue to analyze this data, focusing on those packages where the differences
between the two approaches are large.

       The simulation results may present opportunities to improve the fidelity of the
lumped-parameter approach by identifying differences between different platforms or
important vehicle traits (such as displacement-to-weight ratio, e.g.). There might also be
opportunity to infer (through detailed analysis) the individual effectiveness values for some
technologies by comparing and isolating Ricardo package results across different vehicle
platforms.
                                         1-37

-------
Draft Regulatory Impact Analysis
1.6  Using the Lumped-Parameter Technique to Determine Synergies in a
     Technology Application  Flowpath (Identifying "Technology Pairs" to
     account for synergies)

       In order to account for the real world synergies of combining of two or more
technologies, the product of their individual effectiveness values must be adjusted based on
known interactions, as noted above.  When using an approach in which technologies are
added sequentially in a pre-determined application path to each individual vehicle model, as
used in NHTSA's 2006 fuel economy rule for light trucks3, these interactions may be
accounted for by  considering a series of interacting technology pairs. EPA believes that a
lumped parameter approach can be used as a means to estimate and account for synergies for
such a technology application method. When using a sequential technology application
approach which applies more than one technology, it is necessary to separately account for the
interaction of each unique technology pair. Moreover, if the sequential technology
application approach applies a technology that supersedes another, for example, where a
VVLT system is substituted in place of a cylinder deactivation system, its incremental
effectiveness must be reduced by the sum of the synergies of that technology with each
individual technology that was previously applied, regardless of whether any of them have
also been superseded. Figure 1-3 below provides an example of how technology pairs are
identified for a specific technology application path similar to one used by NHTSA. In this
example, an interaction is identified between each of the engine technologies (except GDI)
with each of the transmission technologies. So,  in this example, were the model to couple a
turbocharged and downsized GDI engine with a 6-speed transmission, it would apply a series
of many synergy  pairs to the combined individual effectiveness values  to arrive at the overall
effectiveness.
              Engine Technology
Trans Technology
                 (Lines indicate potential synergies)

        Figure 1-3 Illustration of technology pairings for a specific technology application path
                                           1-38

-------
                                            Technology Packages, Cost and Effectiveness
       References

       All references can be found in the EPA DOCKET:  EPA-HQ-OAR-2009-0472.
1 National Energy Modeling System, Energy Information Administration, U. S. Dept of Energy.

2 "A Study of Potential Effectiveness of Carbon Dioxide Reducing Vehicle Technologies," EPA Report No.
EPA420-R-08-004, available in the EPA docket EPA-HQ-OAR-2009-0472 and on the Internet at
http://www.epa.gov/otaq/technology/420r08004a.pdf.
                                           1-39

-------
Draft Regulatory Impact Analysis
CHAPTER 2: Air Conditioning
2.1 Overview of Air Conditioning Impacts and Technologies

       Over 95% of the new cars and light trucks in the United States are equipped with air
conditioner (MAC) systems. In the 1970's and 1980's, air conditioner systems were an
optional (luxury) feature, but it now comes standard on almost all new vehicle models. The
Mobile Air Conditioner (A/C) system is a unique and distinct technology on the automobile.
It is different from the other technologies described in Chapter 3 of the joint Technical
Support Document (TSD) in several ways.  First, most of the technologies described in the
joint TSD directly affect the efficiency of the engine, transmission, and vehicle systems.  As
such, these systems are almost always active while the vehicle is moving down the road or
being tested on a dynamometer for the fuel economy and emissions test drive cycles.  A/C on
the other hand, is a parasitic load on the engine that only burdens the engine when the vehicle
occupants demand it.  Since it is not tested as a normal part of the fuel economy and
emissions test  drive cycles,  it is referred to as an "off-cycle" effect. There are many other off-
cycle loads that can be switched on by the occupant that affect the  engine;  these include
lights,  wipers,  stereo systems, electrical defroster/defogger, heated seats, power windows, etc.
However, these electrical loads individually amount to a very small effect on the engine
(although together they can be significant). The A/C system (by itself) adds a significantly
higher load on the engine as described later in this chapter. Secondly, present A/C systems
leak a powerful greenhouse gas directly into the air - even when the vehicle is not in
operation.  No other vehicle system does this.  Because of these factors, a distinct approach to
control of MAC  systems is justified, and a separate technical discussion is also warranted.

       As just mentioned above, there are two mechanisms by which A/C systems contribute
to the emissions  of greenhouse gases. The first is through direct leakage of the refrigerant
into the air. The hydrofluorocarbon refrigerant compound  currently used in all recent model
year vehicles isR134a(also known as 1,1,1,2-Tetrafluoroethane, orHFC-134a).  Based on
the higher global warming potential of HFCs, a small leakage of the refrigerant has a greater
global warming impact than a similar amount of emissions of some other mobile source
GHGs. R134a has a global  warming potential (GWP)  of 1430.A This means that 1 gram of
R134a has the  equivalent global warming potential of  1,430 grams of CO2 (which has a GWP
A The global warming potentials (GWP) used in the NPRM analysis are consistent with Intergovernmental Panel
on Climate Change (IPCC) Fourth Assessment Report (AR4). At this time, the IPCC Second Assessment Report
(SAR) global warming potential values have been agreed upon as the official U.S. framework for addressing
climate change. The IPCC SAR GWP values are used in the official U.S. greenhouse gas inventory submission
to the United Nations climate change framework. When inventories are recalculated for the final rule, changes
in GWP used may lead to adjustments.
                                         2-2

-------
                                                                    Air Conditioning

of I).1  In order for the A/C system to take advantage of the refrigerant's thermodynamic
properties and to exchange heat properly, the system must be kept at high pressures even
when not in operation. Typical static pressures can range from 50-80 psi depending on the
temperature, and during operation, these pressures can get to several hundred psi.  At these
pressures leakage can occur through a variety of mechanisms. The refrigerant can leak slowly
through seals, gaskets, and even small failures in the containment of the refrigerant.  The rate
of leakage may also increase over the course of normal wear and tear on the system. Leakage
may also increase more quickly through rapid component deterioration such as during vehicle
accidents, maintenance or end-of-life vehicle scrappage (especially when refrigerant capture
and recycling programs are less efficient). Small amounts of leakage can also occur
continuously even in extremely "leak-tight" systems by permeating through hose membranes.
This last mechanism is not dissimilar to fuel permeation through porous fuel lines.
Manufacturers may be able to reduce these leakage emissions through the implementation of
technologies such as leak-tight, non-porous, durable components.  The global warming impact
of leakage emissions also can be addressed by using alternative refrigerants with lower global
warming potential.  Refrigerant emissions can also occur during maintenance and at the end
of the vehicle's life (as well as emissions during the initial charging of the system with
refrigerant), and these emissions are already addressed by the CAA Title VI stratospheric
ozone program, as described below.

       The second mechanism by which vehicle A/C systems contribute to GHG  emissions is
through the consumption of additional fuel required to provide power to the A/C system and
from carrying around the weight of the A/C system hardware year-round.  The additional fuel
required to run the system is converted into  CCh by the engine during combustion. These
increased emissions due  to A/C operation can be reduced by increasing the overall efficiency
of the vehicle's A/C system, as described below. EPA will  not be addressing modifications to
the excess weight of the A/C system, since the incremental increase in CCh emissions and fuel
consumption due to carrying the A/C system is directly measured during the normal federal
test procedure, and is thus already subject to the normal control program.

       EPA's analysis indicates that together, these (A/C related) emissions account for about
9% of the greenhouse gas emissions from cars and light trucks.  In this document, EPA will
separate the discussion of these two categories of A/C-related emissions because of the
fundamental differences  in the emission mechanisms and the methods of emission control.
Refrigerant leakage control is akin in many respects to past  EPA fuel evaporation control
programs (in that containment of a fluid is the key  feature),  while efficiency improvements
are more similar to  the vehicle-based control of CCh set out in the joint TSD (in that they
would be achieved through specific hardware and controls).

       EPA recognizes that California and the European Union also believe that A/C related
emissions account for a significant part of greenhouse gas emissions.  Both California and the
European Union have either proposed or discussed programs to limit GHGs from A/C
systems. EPA has evaluated these programs and this document discusses some similar
features and others  that emphasize additional emission reduction mechanisms.
                                         2-3

-------
Draft Regulatory Impact Analysis

2.2 Air Conditioner Leakage

2.2.1 Impacts of Refrigerant Leakage on Greenhouse Gas Emissions

There have been several studies in the literature which have attempted to quantify the
emissions (and impact) of air conditioner HFC emissions from light duty vehicles. In this
section, several of these studies are discussed.

    2.2.1.1  In-Use Leakage Rates

       Based on measurements from 300 European vehicles (collected in 2002 and 2003),
Schwarz and Harnisch estimate that the average HFC direct leakage rate from modern A/C
systems was estimated to be 53 g/yr.2 This corresponds to a leakage rate of 6.9% per year.
This was estimated by extracting the refrigerant from recruited vehicles and comparing the
amount extracted to the  amount originally filled (as per the vehicle specifications).  The fleet
and size of vehicles differs from Europe and the United States, therefore it is conceivable that
vehicles in the United States could have a different leakage rate. The authors measured the
average charge of refrigerant at initial fill to be about 747 grams (it is somewhat higher in the
U.S. at 770g), and that the smaller cars (684 gram charge) emitted less than the higher charge
vehicles (883  gram charge).  Moreover, due to the climate differences, the A/C usage patterns
also vary between the two continents, which may influence leakage rates.

       Vincent et al., from the California Air Resources Board estimated the in-use
refrigerant leakage rate to be 80 g/yr.3 This is based on consumption of refrigerant in
commercial fleets, surveys of vehicle owners and technicians. The study assumed an average
A/C charge size of 950 grams and a recharge  rate of 1 in 16 years (lifetime).  The recharges
occurred when the system was 52% empty and the fraction recovered at end-of-life was 8.5%.

    2.2.1.2  Emission  Inventory

       The EPA publishes an inventory of greenhouse gases and sinks on an annual basis.
The refrigerant emissions numbers that are used in the present analysis are from the Vintaging
model, which is used to generate the emissions included in this  EPA inventory source.  The
HFC refrigerant emissions from light duty vehicle A/C systems was estimated to be 61.8 Tg
CO2 equivalent in 2005 by the Vintaging model.4'8

       In 2005, refrigerant leakage accounted for about 5.1% of total greenhouse gases from
light duty sources.  The  following table shows the breakdown of greenhouse gases as broken
down by the different emissions processes in  2005. The baseline tailpipe  CCh, N2O and CH4
emissions are from MOVES, the refrigerant emissions are from the Vintaging model, and the
A/C CO2 emissions are from EPA and the National Renewable  Energy Laboratory (NREL) as
described below.
B EPA reported the MVAC emissions at 56.6 Tg CO2 EQ, using a GWP of 1300. This number has been adjusted
using a GWP of 1430.
                                        2-4

-------
                                                                     Air Conditioning

     Table 2-1. CO2 equivalent emissions from light duty vehicles broken up by source or process.
Emissions source or process
Tailpipe CO2 (w/o A/C)
CO2 from A/C
HFC-134a (Leakage)
N2O
CH4
Total
Tg CO2 (equivalent)
1,076
47.2
61.8
28.2
1.9
1,215
Percentage of total
88.6%
3.9%
5.1%
2.3%
0.2%

       From a vehicle standpoint, the Vintaging model assumes that 42% of the refrigerant
emissions are due to direct leakage (or "regular" emissions), 49% for service and maintenance
(or "irregular" emissions), and 9% occurs at disposal or end-of-life as shown in the following
table. These are based on assumptions of the average amount of chemical leaked by a vehicle
every year, how much is lost during service of a vehicle (from professional service center and
do-it-yourself practices), and the amount lost at disposal. These numbers vary  somewhat over
time based on the characteristics (e.g. average charge size and leakage rate) of each "vintage"
of A/C system, assumptions of how new A/C systems enter the market, and the number of
vehicles disposed of in any given year.

 Table 2-2: Light duty vehicle HFC-134a emissions in 2005 from Vintaging model. HFC emissions can be
                   converted to CO2 equivalent by multiplying by 1430 GWP.
Emission Process
Leakage
Maintenance/servicing
Disposal/end-of-life
Total
HFC emissions (metric tons)
18,151
21,176
3,890
43,217
Fraction of total
0.42
0.49
0.09
1.0
2.2.2 A/C Leakage Credit

       The level to which each technology can reduce leakage can be calculated using the
SAE Surface Vehicle Standard J2727 - HFC-134a Mobile Air Conditioning System
Refrigerant Emission Chart.  This industry standard was developed by SAE and the
cooperative industry and government IMAC (Improved Mobile Air Conditioning) program
using industry experience, laboratory testing of components and systems, and field data to
establish a method for calculating leakage.  With refrigerant leakage rates as low as 10 g/yr, it
would be exceedingly difficult to measure such low levels in a test chamber (or shed).  Since
the J2727 method has been correlated to "mini-shed" results (where select components are
tested in a small chamber, simulating real-world driving cycles), the EPA considers this
method to be an appropriate surrogate for vehicle testing of leakage.  It is also referenced by
the California Air Resources Board in their Environmental Performance Label regulation and
the State of Minnesota in their GHG reporting regulation.5'6

    2.2.2.1  Why Is EPA Relying on a Design-Based Rule?

       As with any design-based rule, it is possible to achieve compliance by simply
selecting the minimum number of design attributes needed to meet a particular threshold or
                                         2-5

-------
Draft Regulatory Impact Analysis

standard. Whether a design-based approach is used for emissions compliance or earning
voluntary GHG credits, manufacturers will rightly choose the combination of design attributes
which yield the maximum benefit at the lowest cost. However, there is a risk that some
manufacturers may select poor quality, cheap parts, or implement the changes poorly,
resulting in vehicles which ostensibly meet the rule's provisions, but in practice, fail to
achieve their stated benefits. However, EPA believes that the market-driven incentive of
assuring customer satisfaction will drive manufacturers to design A/C systems that perform as
promised, and never need to be recharged. Also, it should be noted that the relative leakage
rates assigned to various components, materials, and technologies in SAE J2727 are based on
(and correlated to) actual leakage rates, as measured in bench- and field-test studies of
vehicles and components.

       In the case of refrigerant leakage, it would be very costly and burdensome to design,
develop, and implement a test procedure and facility for measuring refrigerant leakage on
each and every vehicle type a manufacturer produces.  With leakage levels on many new
vehicles expected to be as low as 9 g/yr (0.001 g/hr), it would be difficult to accurately
measure the actual leakage rate.  Even if it were possible to build a suitable facility capable of
accurately measuring very low levels of refrigerant leakage, such a facility would still not
exercise the A/C system across its normal range of operation, both in terms of engine and
vehicle  speeds as well as ambient conditions (e.g. under high compressor load, leakage past
the compressor shaft seal on a running system can be 20 times higher than the static leakage
level).7  In addition, it is very likely that any performance-based test would become obsolete in
the timeframe of this rulemaking, as low-GWP refrigerants are likely to be adopted by
manufacturers.

       In the absence  of a vehicle-level performance test to measure the how a particular A/C
system design functions (and the difficulty in creating  such a test), EPA is proposing to rely
upon the best available design metrics for quantifying  system performance.  EPA believes that
the SAE J2727 method as an appropriate method for quantifying the expected yearly
refrigerant  leakage rate from A/C systems.

    2.2.2.2  How Are Credits  Calculated?

       The A/C credit available to manufacturers will be calculated based on how much a
particular vehicle's annual leakage value is reduced against the average new vehicle, and will
be calculated using a method drawn directly from the SAE J2727 approach. By scoring the
minimum leakage rate possible on the J2727 components enumerated in the proposed rule
(expressed  in the proposed rule as a measure of annual leakage), one earns the maximum A/C
credit (on a gram per mile basis).

       The A/C credit available to manufacturers will be calculated based on the reduction to
a vehicle's  yearly leakage rate, using the following equation:
                                         2-6

-------
                                                                    Air Conditioning
Equation 1 - Credit Equation

A/C Credit = (MaxCredit) * [ 1 - (§86.166-12 Score/Avglmpactc) * (GWPRefrigerant/1430)]
There are four significant terms to the credit equation.  Each is briefly summarized below, and
is then explained more thoroughly in the following sections. Please note that the values of
many of these terms change depending on whether HFC-134a or an alternative refrigerant are
used. The values are shown in Table 2-3, and are documented in the following sections.

   •  "MaxCredit" is a term for the maximum amount of credit entered into the equation
       before constraints are applied to terms. The maximum credits that could be earned by
       a manufacturer is limited by the choice of refrigerant and by assumptions regarding
       maximum achievable leakage reductions.

   •  " Score/ Avglmpact" is the leakage score of the A/C system as measured according to
       the §86.166-12 calculation in units of g/yr, where the minimum score which is deemed
       feasible is fixed.

   •  "Avglmpact" is the annual average impact of A/C leakage.

   •  "GWPRefrigerant" is the global warming potential for direct radiative forcing of the
       refrigerant as defined by EPA (or IPCC).
                     Table 2-3: Components of the A/C Credit Calculation


MaxCredit equation input (grams /mile CCh EQ)
A/C credit maximum (grams /mile CO2 EQ)
§86.166-12 Score Avglmpact (grams / HFC year)
Avg Impact (grams / HFC year)
HFC-134a
Cars
12.6
6.3
8.3
16.6
Trucks
15.6
7.8
10.4
20.7
Alternative
Refrigerant
Cars
13.8
13.8
8.3
16.6
Trucks
17.2
17.2
10.4
20.7
2.2.2.2.1 Max Credit Term

       In order to determine the maximum possible credit on a gram per mile basis, it was
necessary to determine the projected real world HFC emissions per mile in 2016.  Because
HFC is a leakage type emission, it is largely disconnected from vehicle miles traveled
 Proposed section 86.166-12 sets out the individual component leakage values based on the SAE value.
                                         2-7

-------
Draft Regulatory Impact Analysis

(VMT).D  Consequently, the total HFC inventory in 2016 was calculated, and then calculated
the relevant VMT. The quotient of these two terms is the HFC contribution per mile.

       Consistent with the methodology presented in DRIA chapter 5, the HFC emission
inventories were estimated from a number of existing data sources. The per-vehicle per-year
HFC emission of the current (reference) vehicle fleet was determined using averaged 2005
and 2006 registration data from the Transportation Energy Databook (TEDB) and 2005 and
2006 mobile HFC leakage estimates from the EPA Emissions and Sinks report described
above.8'9 The per-vehicle per-year emission rates were then adjusted to account for the new
definitions of car and truck classes (described in preamble section I), by increasing the car
contribution proportionally by the percentage of former trucks that are reclassified as cars.
This inventory calculation assumes that the leakage rates and charge sizes of future fleets are
equivalent to the fleet present in the 2005/2006 reference years. Preliminary EPA analysis
indicates that this may increasingly overstate the future HFC inventory, as charge sizes are
decreasing.

       The per-vehicle per-year average emission rate was then scaled by the projected
vehicle fleet in each future year (using the fleet predicted in the emissions analysis) to
estimate the HFC emission inventory if no controls were enacted on the fleet. After dividing
the 2016 inventory by total predicted VMT in 2016, an average per mile HFC emission rate
("base rate") was obtained.

       The base rate is an average in-use number, which includes both old vehicles with
significant leakage, as well as newer vehicles with very little leakage.  The new vehicle
leakage rate is discussed in section 2.2.2.2.2, while deterioration is discussed in section 2.2.5.

   •    Max Credit with Conventional Refrigerant (HFC-134a)
       Two adjustments were made to the base rate in order to calculate the Maximum HFC
       credit with conventional refrigerant.  First, EPA has determined that 50% leakage
       prevention is  the maximum potentially feasible prevention rate in the 2012-2016
       timeframe (section 2.2.3).  Some leaks will occur and are expected, regardless of
       prevention efforts.  The accuracy of the J2727 approach (as expressed in proposed
       §86.112), as a design based test, decreases as the amount of expected leakage
       diminishes. 50% of the base rate is therefore proposed to be set as the maximum
       potential leakage credit for improvements to HFC leakage using conventional
       refrigerant.

       Second, EPA expects that improvements to conventional refrigerant systems will
       affect both leakage and service emissions, but will not affect end of life emissions.
       EPA expects that reductions in the leakage rate from A/C systems will result in fewer
       visits for maintenance and recharges. This will have the side benefit of reducing the
       emissions leftover from can heels (leftover in the recharge cans) and the other releases
D In short, leakage emissions occur even while the car is parked, so the connection to a gram/mile credit is not
straightforward. However, HFC emissions must be converted to a gram/mile basis in order to create a relevant
credit.

-------
                                                                    Air Conditioning

       that occur during maintenance. However, as disposal/end of life emissions will be
       unaffected by the leakage improvements (and also are subject to control under the
       rules implementing Title VI of the CAA), the base rate was decreased by a further 9%
       (Table 2-2).

   •    Max Credit with Alternative Refrigerant
       Emission reductions greater than 50% are possible with alternative refrigerants.  As an
       example, if a refrigerant with a GWP of 0 were used, it would be possible to eliminate
       all refrigerant GHG emissions. In addition, for alternative refrigerants, the EPA
       believes that vehicles with reduced GWP refrigerants should get credit for end of life
       emission reductions. Thus, the maximum credit with alternative refrigerant is about
       9% higher than twice the maximum leakage reduction.

       A final adjustment was made to each credit to account for the difference between real-
world HFC emissions and test-cycle CCh emissions. It has been shown that the tests currently
used for CAFE certification represents an approximately 20% gap from real world fuel
consumption and the resulting CCh emissions.10  Because the credits from direct a/c
improvements are taken from a real world source, and are being traded for an increase in fuel
consumption due to increased CCh emissions, the credit was multiplied by 0.8 to maintain
environmental neutrality (Table 2-4).
          Table 2-4 HFC Credit Calculation for Cars and Trucks based on a GWP of 1430









Car
Truck
Total
HFC
Inventory
(MMT
C02EQ)





27.4
30.4
57.8
VMT
(Billions
of Miles)






1,580
1,392
2,972
Total HFC
EmissionsPer
Mile
(C02EQ
Gram/mile)




17.2
21.5
18.6
HFC
Leakage and
Service
EmissionsPer
Mile
(CO2EQ
Gram/mile)


15.5
19.6
16.9
Maximum
Credit w/
alternative
refrigerant
(Adjusted
for On-
road gap &
including
end of life)
13.8
17.2
14.9
Maximum
Credit w/o
alternative
refrigerant
(50% of
Adjusted
HFC&
excluding
end of life)
6.3
7.8
6.8
2.2.2.2.2 Proposed section 86.166-12, implementing the J2727 Score Term

       The J2727 score is the SAE J2727 yearly leakage estimate of the A/C system as
calculated according to the J2727 procedure. The minimum score for cars and trucks is a
fixed value, and the section below describes the derivation of the minimum leakage scores
that can be  achieved using the J2727 procedure.

       In contrast to the studies discussed in section 2.2.1.1 which discussed the HFC
emission rate of the in-use fleet (which includes vehicles at all stages of life), the SAE J2727
estimates leakage from new vehicles. In the development of J2727, two relevant studies were
                                         2-9

-------
Draft Regulatory Impact Analysis

assessed to quantify new vehicle emission rates. In the first study, measurements from
relatively new (properly functioning and manufactured) Japanese-market vehicles were
collected.  This study was based on 78 in-use vehicles (56 single evap, 22 dual evap) from 7
Japanese auto makers driven in Tokyo and Nagoya from April, 2004 to December, 2005.  The
study also measured a higher emissions level of 16 g/yr for 26 vehicles in a hotter climate
(Okinawa). This study indicated the leakage rate to be close to 8.6 g/yr for single evaporator
systems and 13.3 g/yr for dual evaporator systems.11 A weighted (test) average gives 9.9 g/yr.
In the second study, emissions were measured on European-market vehicles up to seven years
age driven from November,  2002 to January, 2003.12 The European vehicle emission rates
were slightly higher than the Japanese fleet, but overall, they were consistent. The average
emission rate from this analysis is 17.0 g/yr with a standard deviation of 4.4 g/yr. European
vehicles, because they have  smaller charge sizes, likely understate the leakage rate relative to
the United States.  To these  emission rates, the J2727 authors added a factor to account for
occasional defective parts and/or improper assembly and to calibrate the result of the SAE
J2727 calculation with the leakage measured in the vehicle and component leakage studies.

       We adjust this rate up slightly by a factor proportional to the average European
refrigerant charge to the average United States charge (i.e. 770/747 from the Vintaging model
and Schwarz studies respectively). The newer vehicle emission rate is thus  18 g/yr for the
average newer vehicle emissions. This number is a combined car and truck number, and
although based on the limited data, it was not possible to separate them.

       To derive the minimum score, the 18 gram per year rate was used as a ratio to convert
the gram per mile emission impact into a new vehicle gram per year for the test. The car or
truck direct a/c emission factor (gram per mile) was divided by the average emission factor
(gram per mile) and then multiplied by the new vehicle average  leakage rate (gram per year)
Equation 2 - J2727 Minimum Score
       J2727 Minumum Score = Car or truck average pre control emissions (gram per
       mile)/ Fleet average pre-control emissions (grams per mile) x New vehicle annual
       leakage rate (grams per year) x Minimum Fraction
       By applying this equation, the minimum J2727 score is fixed at 8.3 g/yr for cars and
10.4 g/yr for trucks. This corresponds to a total fleet average of 18 grams per year, with a
maximum reduction fraction of 50%

       The GWP Refrigerant term in Equation 1 allows for the accounting of refrigerants
with lower GWP (so that this term can be as low as zero in the equation), which is why the
same minimum score is kept regardless of refrigerant used.

       It is technically feasible for the J2727 Minimum score to be less than the values
presented in the table. But this will usually require the use of an electric compressor (see
below for technology description), which the EPA does not expect to see with high
                                        2-10

-------
                                                                    Air Conditioning

penetrations within the 2012-2016 timeframe, as this technology is likely to accompany
hybrid vehicle and stop-start technologies, and not conventional vehicles.

2.2.2.2.3 Avglmpact Term

       Avglmpact is the average annual impact of A/C leakage, which is 16.6 and 20.7 g/yr
for cars and trucks respectively. This was derived using Equation 2, but by setting the
minimum fraction to one.

2.2.2.2.4 GWPRefrigerant Term

       This term is relates to the global warming potential (GWP) of the refrigerant as
documented by EPA. A full discussion of GWP and its derivation is too lengthy for this
space, but can be found in many EPA documents.13 This term is used to correct for
refrigerants with global warming potentials that differ from HFC-134a. As just explained,
this term accounts for the GWP of any refrigerant used, and can be as low as zero.

2.2.3 Technologies That Reduce Refrigerant Leakage and their Effectiveness

       In this section, the baseline technologies which were used in the EPA's analysis of
refrigerant leakage are described as well as the effectiveness of the leakage-reducing
technologies that are believed will be available to manufacturers in the 2012-to-2016
timeframe of this proposed rulemaking.  An EPA analysis to determine a baseline leakage
emission rate was conducted in the 2006-to-2007 timeframe, and at that time, it was estimated
that the A/C system in new vehicles would leak refrigerant at an average rate of 18 g/yr,
which represents the types of A/C components and technologies currently in use.  EPA
believes, through utilization of the leakage-reducing technologies described below, that it will
be possible for manufacturers to reduce refrigerant leakage 50%, relative to the 18 g/yr
baseline level.14 EPA also believes that all of these leakage-reducing technologies are
currently available, and that many manufacturers have already begun using them to improve
system reliability and in anticipation of the State of California's Environmental Performance
Label regulations  and the State of Minnesota's reporting requirements for High Global
Warming Potential Gases.

       In describing the technologies below, only the relative effectiveness figures are
presented, as the individual piece costs are not known. The EPA only has costs of complete
systems based on the literature, and the individual technologies are described below.

    2.2.3.1  Baseline Technologies

       The baseline technologies assumed for A/C systems which have an average annual
leak rate of 18 g/yr are common to many mass-produced vehicles in the United States. In
these mass-produced vehicles, the need to maintain A/C system integrity  (and the need to
avoid the customer inconvenience of having their A/C system serviced due to loss of
refrigerant) is often balanced against the cost of the individual A/C components. For
manufacturers seeking improved system reliabilty, components and technologies which
reduce leakage (and possibly increase cost) are selected, whereas other manufacturers may
                                        2-11

-------
Draft Regulatory Impact Analysis

choose to emphasize lower system cost over reliabilty, and choose components or
technologies prone to increased leakage.  In the absence of standards or credits concerning
refrigerant leakage, it is the market forces of cost and reliability which determine the
technology a manufacturer chooses.  In EPA's baseline scenario, the following assumptions
were made  concerning the definition of a baseline A/C system:

                 -   all flexible hose material is rubber, without leakage-reducing barriers or
                    veneers, of approximately 650 mm in length for both the high and low
                    pressure lines
                 -   all system fittings and connections are sealed with a single o-rings
                 -   the compressor shaft seal is a single-lip design
                 -   one access port each on the high and low pressure lines
                 -   two of the following components: pressure switch, pressure relief
                    valves, or pressure transducer
                 -   one thermostatic expansion valve (TXV)

       The design assumptions of EPA baseline scenario are also similar to the sample
worksheet included in SAE's surface vehicle standard J2727 - HFC-134a Mobile Air
Conditioning System Refrigerant Emission Chart.15 In the J2727 emission chart, it is the
baseline technologies which are assigned the highest leakage rates, and the inclusion of
improved components and technologies in an A/C system will reduce this annual leakage rate,
as a function of their effectiveness relative to the baseline. EPA considers these 'baseline'
technologies to be representative of recent model year vehicles, which, on average, can
experience  a refrigerant loss of 18 g/yr. However, depending on the design of a particular
vehicle's A/C system (e.g. materials, length of flexible hoses, number of fittings and adaptor
plates, etc.), it is possible to achieve a leakage score much higher (i.e. worse) than 18 g/yr.
According to manufacturer data submitted to the State of Minnesota, 19% of 2009 model year
vehicles have a J2727 refrigerant score greater than 18 g/yr, with the highest-scoring vehicle
reporting a leakage rate of 30.1 g/yr.16 The average leakage was found to be 15.1 g/yr,
though this value  is not sales weighted.

    2.2.3.2  Flexible Hoses

       The flexible hoses on an automotive A/C system are needed to isolate the system from
engine vibration and to  allow for the engine to roll within its mounts as the vehicle accelerates
and decelerates.  Since the compressor is typically mounted to the  engine, the lines going to-
and-from the compressor (i.e. the suction and pressure lines) must  be flexible,  or unwanted
vibration would be transferred to the body of the vehicle (or other components), and excessive
strain on the lines would result. It has been industry practice for many years to manufacture
these hoses from rubber, which is relatively inexpensive and durable.  However, rubber hoses
are not impermeable, and refrigerant gases will eventually migrate into the atmosphere. To
reduce permeation, two alternative hose material can be specified.  The first material, is
known as a standard 'veneer' (or 'barrier') hose, where a polyamide (polymer) layer - which
has lower permeability than rubber - is encased by a rubber hose. The barrier hose is similar
to a veneer hose, except that an additional layer of rubber is added inside the polyamide layer,
creating three-layer hose (rubber-polyamide-rubber). The second material is known as 'ultra-
                                         2-12

-------
                                                                     Air Conditioning
low permeation', and can be used in a veneer or barrier hose design. This ultra-low
permeation hose is the most effective at reducing permeation, followed by the standard veneer
or barrier hose. Permeation is most prevalent during high pressure conditions, thus it is even
more important that these low permeable hoses are employed on the high pressure side, more
so than on the low pressure side.  EPA expects that many manufacturers will begin using
these technologies (and many have already begun doing so) to reduce refrigerant leakage.

       According to J2727, standard barrier veneer hoses have 25% the permeation rate of
rubber hose, and ultra low permeable barrier veneer hoses have 10% the permeation rate (as
compared to a standard baseline rubber hose of the same length and diameter).

     2.2.3.3   System Fittings and Connections

       Within an automotive A/C system and the various components it contains (e.g.
expansion valves, hoses, rigid lines, compressors, accumulators, heat exchangers, etc.), it is
necessary that there be an interface, or connection, between these components. These
interfaces may exists for design, manufacturing, assembly, or serviceability reasons, but all
A/C systems have them to some degree, and each interface is a potential path for refrigerant
leakage to the atmosphere. In SAE J2727 emission chart, these interfaces are described as
fittings and connections, and each type of fitting or connection type is assigned an emission
value based on its leakage potential; with a single o-ring (the baseline technology) having the
highest leak potential; and a metal gasket having the lowest. In between these two extremes,
a variety of sealing technologies, such as multiple o-rings, seal washers, and seal washers with
o-rings, are available to manufacturers for the purpose of reducing leakage. It is expected that
manufacturers will choose from among these sealing technology options to create an A/C
system which offers the best cost-vs-leakage rate trade-off for their products.

       The relative effectiveness of the fitting and connector technology is presented in Table
2-5.  For example, the relative leakage factor of 125 for the baseline single O-ring is 125
times more "leaky" than the best technology - the metal gasket.

       Table 2-5 :  Effectiveness of Fitting and Connector Technology
Fitting or Connector
Single O-ring
Single Captured O-ring
Multiple O-ring
Seal Washer
Seal Washer with O-ring
Metal Gasket
Relative
Leakage
125
75
50
10
5
1
                                         2-13

-------
Draft Regulatory Impact Analysis
    2.2.3.4  Compressor Shaft Seal

       A major source of refrigerant leakage in automotive A/C systems is the compressor
shaft seal.  This seal is needed to prevent pressurized refrigerant gasses from escaping the
compressor housing. As the load on the A/C system increases, so does the pressure, and the
leakage past the seal increases as well.  In addition, with a belt-driven A/C compressor, a side
load is placed on the compressor shaft by the belt, which can cause the shaft to deflect
slightly. The compressor shaft seal must have adequate flexibility to compensate for this
deflection, or movement, of the compressor shaft to ensure that the high-pressure refrigerant
does not leak past the seal lip and into the atmosphere.  When a compressor is static (not
running), not only are the system pressures lower, the only side load on the compressor shaft
is that from tension on the belt, and leakage past the compressor shaft is at a minimum.
However, when the compressor is running, the  system pressure is higher and the side load on
the compressor shaft is higher (i.e.  the side load is proportional to the power required to turn
the compressor shaft) - both of which can increase refrigerant leakage past the compressor
shaft seal. It is estimated that the rate of refrigerant leakage when a compressor is running
can be 20 times that of a static condition.7 Due to the higher leakage rate under running
conditions, SAE J2727 assigns a higher level of impact to the compressor shaft seal. In the
example shown in the August 2008 version of the J2727 document, the compressor is
responsible for 58% of the system refrigerant leakage, and of that 58%, over half of that
leakage is due  to the shaft seal alone (the remainder comes from compressor housing and
adaptor plate seals). To address refrigerant  leakage past the compressor shaft, manufacturers
can use multiple-lip seals in place of the single-lip seals.

2.2.4  Technical Feasibility of Leakage-Reducing Technologies

       EPA believes that the leakage-reducing technologies discussed in the previous
sections are available to manufacturers  today, are relatively low in cost, and that their
feasibility and  effectiveness have been demonstrated by the SAE IMAC teams.  EPA also
believes - as has been demonstrated in the J2727 calculations submitted by manufacturers to
the State of Minnesota - that  reductions in leakage from 18 g/yr to 9 g/yr are possible (e.g. the
2009 Saturn Vue has a reported leakage score of 8.5 g/yr). In addition to earning credit for
reduced refrigerant leakage, some manufacturers  may, within the timeframe of this
rulemaking, choose to introduce alternative  refrigerant systems, such as HFO-1234yf.

2.2.5  Deterioration of Leakage Controls in A/C Systems

       In order to determine the cost savings from the improvements to the leakage system, it
is necessary to project the point at which the vehicle will require servicing and an additional
refrigerant charge.

       There are two mechanisms  of deterioration that are modeled: the normal deterioration
that results in increasing leakage and the "avoidable" deterioration of the condenser &
compressor components.  This model is developed to help us estimate the costs of the A/C
                                        2-14

-------
                                                                       Air Conditioning

reductions.  It is especially needed to determine the period over which the discounted cost
savings should be applied.E

       Normal deterioration occurs throughout all components of the A/C system. Hoses,
fittings, compressors, etc all wear with age and exposure to heat (temperature changes),
vibration, and the elements. It is assumed that the system deterioration rates decrease
(proportionally) as the base leakage rates are decreased with the use of improved parts and
components. The base deterioration rate is modeled as a linear function, such that the (new
vehicle) leakage rate is 18 g/yr at age zero and 59 g/yr at the "average" age of 5 years old.
The 18 gram leakage rate for new vehicles has been documented in section 2.2.2,  while the 59
gram mid-life leakage rate is drawn from the Vintaging model and is documented below.

       The Vintaging model assumes a constant leakage + servicing emission rate of 18% per
year for modern vehicles running with HFC-134a refrigerant.  As the emission rates do not
change by age invintaging, the emission rate is the average rate of loss over the vehicle's life.

        Applying the percentages in Table 2-2, this corresponds to a leakage rate of 7.6% (59
grams) per year and a servicing loss rate of 8.8% (68 grams) per year averaged over the
vehicle's life. The model assumes an average refrigerant charge of 770 grams for vehicles
sold in 2002 or later and does not currently assume that these charge sizes will change in the
future; however, the model may be updated as new information becomes available. The
resulting vehicle emission rates are presented in Table 2-6.

           Table 2-6: Annual in-use vehicle HFC-134a emission rate from Vintaging model.
Emission Process
Leakage
Servicing/maintenance
Leak rate (%/year)
7.6%
8.8%
Leak rate (g/year)
59
68
       The average leakage emissions rate of 59-68 g/yr is higher with Schwarz's European2
study and lower than CARB's study,3 and thus is within the range of results in the literature.
         Air conditioning leakage controls are the only technology in this proposed rule that has an
assumed deterioration that affects the effectiveness of the technology.  This is partly because sufficient
data is not available for many of the technologies in chapter 3 of the TSD. Moreover, it is not
expected that deterioration of powertrain technologies will lead to emissions increases on the scale of
those seen when criteria pollutant technologies deteriorate. The deterioration from the latter can
increase emissions by factors of 10 or even 100 or more.  Similarly, air conditioning leakage
technologies can and do deteriorate, contributing to significantly higher emissions over time. For this
reason, a deterioration model is proposed below.  This model only applies for leakage, and not for
indirect CO2 (tailpipe) emissions due to A/C. For the latter, a partly functioning system may lead to
somewhat higher emissions, but when it finally fails, it is one of the few technologies where the
emissions are no longer relevant, i.e. an A/C system that no longer functions, no longer emits indirect
emissions.
                                          2-15

-------
Draft Regulatory Impact Analysis
       This model is presented in Figure 2-1 with the assumption that the average vehicle
(A/C system) last about 10 years. Technically, the assumption is that the A/C system lasts 10
years and not the vehicle per se. Inherent in this assumption is that the vehicle owner will not
repair the A/C system on an older vehicle due to the expensive nature of most A/C repairs late
in life relative to the value of the vehicle.  It is also assumed that the refrigerant requires a
recharge when the state of charge reaches 50% for the analysis in this section. This
deterioration/leakage model approach will be used later to estimate the cost of maintenance
savings due to low leak technologies (from refills) as well as the benefits of leakage controls.
             120
             100
              80
o
'in
•|   60
LU
u
              40
          n
          v
                          -•—leakage
                          -B—remaining charge
                                                             -  700 —
                                                             -600 5
                                                             -  500
                                                 recharge
                                                 required
   800
                                                             -  300
                                                             -200
                                                             -- 100 o
                               4       6      8
                                  Age (years)
                                           10
12
                      Figure 2-1. Deterioration rate of refrigerant leakage.

       Figure 2-2 shows how the leakage rates vary with age as the initial leakage rates are
decreased to meet new proposed standards (with improved components and parts). The
deterioration lines of the lower leakage rates were determined by applying the appropriate
ratio to the 17 g/yr base deterioration rate. Figure 2-3 shows the refrigerant remaining, which
includes a line indicating when a recharge is required (50% charge remaining out of an initial
charge of 770g). So a typical vehicle meeting a leakage score of 8.5 g/yr (new) will not
require a recharge until it is about 12 years old.
                                         2-16

-------
                                                                  Air Conditioning
190
1 nn
I UU
£ RD
>i OU
2
« en
O) DU
(0
^
55 /in
O> ^tu
on
zu
n

•
• •
• •
•
A ^

• 18 g/yr
• 13. 5 g/yr
A 9 g/yr
• 4.5 g/yr


A
* . '
* - A * . « '
_ A • •
A • »
: . • •
U I I I I I
0 2 4 6 8 10 12 14
Age of Vehicle
Figure 2-2. A/C refrigerant leakage rate for different technologies as vehicles age.
Rnn -,
^-^ ouu

E ynn
g 'uu
S finn
uuu
O)
.= cnn
^ OUU
'<5
E^nn

Qnn
^j OUU
c
5 onn
a ^uu
O)
'c 1 nn
!p I UU
a)
tf 0
1 I : • . .
: : • ' • .
» • A •
A •
* • A
»
• *
* A

refill line *
# 18 g/yr
• 13. 5 g/yr
A 9 g/yr
• 4.5 g/yr
•


A
•
•

W I I I I I
0 2 4 6 8 10
Age (years)
12 14
                                  2-17

-------
Draft Regulatory Impact Analysis

       Figure 2-3.  A/C refrigerant remaining in a typical system as vehicles age and deteriorate.



2.2.6 Other Benefits of improving A/C Leakage Performance

       The EPA is assuming that a reduction in leakage emissions from new vehicles will
also improve the leakage over the lifetime of the vehicle.  There is ample evidence to show
that A/C systems that leak more also have other problems that occur (especially with the
compressor) due to the lack of oil circulating in the system. Thus, it is expected that an A/C
system which utilizes leak-reducing components and technologies should, on average, last
longer than one which does not.

       An European study conducted in 2001 (by Schwarz) found that the condenser is the
component most likely to fail and result in a total leak.17  The study also found that
compressor component was most likely the culprit when other malfunctions were present
(other than total loss). A more recent (and larger) study found that condensers required
replacement at half the rate of a compressor (10% vs 19% of the entire part replacement rate),
and that evaporators and accumulators failed more often.7  The same study also found that
many of the repairs occurred when the vehicles were aged 5-10 years.  Both these studies
indicate that the condenser and compressor are among the major causes of failure in an A/C
system. Leakage reductions in the system are expected to greatly reduce the incidence of
compressor repair, since one of the main root causes of compressor failure is a shortage of
lubricating oil, which originates from a shortage of refrigerant flowing through the system
(and it is a refrigerant-oil mixture which carries lubricating oil to the compressor).18

       Monitoring of refrigerant volume throughout the life of the A/C system may provide
an opportunity to circumvent some previously described failures specifically related to
refrigerant loss. Similar to approaches used today by the engine on-board diagnostic systems
(OBD) to monitor engine  emissions, a monitoring system that informed the vehicle operator
of a low refrigerant level could potentially result in  significant reductions in A/C refrigerant
emissions due to component failure(s) by creating an opportunity  for early repair actions.
While most A/C systems contain sensors capable of detecting the low  refrigerant pressures
which result from significant refrigerant loss, these systems are generally not designed to
inform the vehicle operator of the refrigerant loss, and that further operation of the system in
this state can result in additional component damage (e.g.  compressor  failure).  Electronic
monitoring of the refrigerant may be achieved by using a combination of existing A/C system
sensors and new software designed to detect refrigerant loss before it progresses to a level
where component failure is likely to occur.
                                         2-18

-------
                                                                 Air Conditioning
2.3 COi Emissions due to Air Conditioners

2.3.1  Impact of Air Conditioning Use on Fuel Consumption and CO2 Emissions

      Three studies have been performed in recent years which estimate the impact of A/C
use on the fuel consumption of motor vehicles. In the first study, the National Renewable
Energy Laboratory (NREL) and the Office of Atmospheric Programs (OAP) within EPA have
performed a series of A/C related fuel use studies.19'20 The energy needed to operate the A/C
compressor under a range of load and ambient conditions was based on testing performed by
Delphi, an A/C system supplier.  They used a vehicle simulation model, ADVISOR, to
convert these loads to fuel use over the EPA's FTP test cycle. They developed a personal
"thermal comfort"-based model to predict the percentage of drivers which will turn on their
A/C systems under various ambient conditions. Overall, NREL estimated A/C use to
represent 5.5% of car and light truck fuel consumption in the U.S.

      In the second study, the California Air Resources Board (ARE) estimated the impact
of A/C use on fuel consumption as part of their GHG emission rulemaking.21  The primary
technical analysis utilized by ARE is summarized in a report published by NESCCAF for
ARE. The bulk of the technical work was performed by two contractors: AVL Powertrain
Engineering and Meszler Engineering Services. This work is founded on that performed by
NREL-OAP. Meszler used the same Delphi testing to estimate the load of the A/C
compressor at typical ambient conditions. The impact of this load on onroad fuel
consumption was  estimated using a vehicle simulation model developed by AVL - the
CRUISE model -  which is more sophisticated than ADVISOR. These estimates were made
for both the EPA FTP and HFET test cycles. (This is the combination of test cycle results
used to determine compliance with NHTSA's current CAFE standards.) NREL's thermal
comfort model was used to predict A/C system use in various states and seasons.

      The NESCAFE results were taken from Table 3-1 of their report and are summarized
in Table 2-7.
           22
             Table 2-7: CO2 Emissions Over 55/45 FTP/HFET Tests and From A/C Use (g/mi)

55/45 FTP/HFET
Indirect A/C Fuel Use
Total
Indirect A/C Fuel Use
Small Car
278
16.8
294.8
5.7%
Large Car
329
19.1
348.1
5.5%
Minivan
376
23.5
399.5
5.9%
Small Truck
426
23.5
449.5
5.2%
Large Truck
493
23.5
516.5
4.6%
      NESCAFE estimated that nationwide, the average impact of A/C use on vehicle fuel
consumption ranged from 4.6% for a large truck or SUV, to 5.9% for a minivan.  The total
                                       2-19

-------
Draft Regulatory Impact Analysis

CO2 emissions were determined using a 55%/45% weighting of CCh emissions from EPA
FTP and HFET tests plus A/C fuel use (hereafter referred to simply as FTP/HFET). .For the
purposes of this analysis of A/C system fuel use, the percentage of CCh emissions and fuel
use are equivalent, since the type of fuel being used is always gasoline.F

       In order to compare the NESCCAF and ARE estimates to that of NREL-OAP,
weighting factors for the five vehicle classes were developed. NESCCAF presented sales
percentages for the five vehicle classes in Table 2-1 of their report.22 These are shown below
in Table 2-8.  Since these sales percentages do not sum to 100% (possibly due to round-off or
because some vehicles do not fit into any of the five categories) the percentages were
normalized so that they summed to 100%.  The car and truck categories were then weighted
by their lifetime VMT, normalized to that of cars. G This meant a relative weighting factor for
the three truck categories of 1.11 relative to a factor of 1.0 for cars.  The percentage of
lifetime VMT represented by each vehicle class were then determined. These estimates are
shown on the last line of Table 2-8.

                             Table 2-8: Sales and VMT by Vehicle Class

NESCCAF sales
Normalized
NESCCAF sales
Lifetime VMT
weighting factor
VMT
Small Car
22%
22.4%
1.00
21.2%
Large Car
25%
25.5%
1.00
24.1%
Minivan
7%
7.1%
1.11
7.5%
Small Truck
23%
23.5%
1.11
24.6%
Large Truck
21%
21.4%
1.11
22.5%
       Using the percentages of VMT represented by each vehicle class, the A/C fuel use
impacts of NESCCAF and ARE were weighted and determined that they represent 5.3% and
4.2% of fuel use over the FTP/HFET, respectively, including the A/C fuel use.

       In the final study, EPA evaluated the impact of A/C use on fuel consumption as part of
its recent rulemaking which revised the onroad fuel economy labeling procedures for new
motor vehicles.23 EPA estimated the impact of the A/C compressor on fuel consumption from
vehicle emission measurements taken over its SC03 emissions test. SC03 is a 10 minute test
where the vehicle is operated at city speeds, at 95 degrees F, 40% relative humidity and a
solar load of 850 Watts/m2. In addition, prior to the test, the vehicle has been pre-heated for
F Because NESCCAF estimated A/C fuel use nationwide, while ARB focused on that in California, the
NESCCAF and EPA methodologies and results are coempared below.

G Based on annual mileage per vehicle from the Volpe Model discounted at 7% per year. Discounted lifetime
mileages are 102,838 for cars and 114,350 for trucks.
                                         2-20

-------
                                                                   Air Conditioning

10 minutes under these conditions, so the interior cabin starts the test at an elevated
temperature. Testing of 500 late model vehicles over both the FTP and SC03 test cycles
indicated that fuel consumption was 27% higher on the SC03 test than over a combination of
Bag 2 and Bag 3 fuel consumption designed to match the vehicle load of the SC03 test. EPA
assumed that the A/C compressor was engaged  100% of the time over SC03 due to the high
ambient temperature, short duration and vehicle pre-heating test conditions.

       EPA does not measure A/C emissions at highway speeds. Thus, this impact had to be
estimated based on the city-like SC03 test.  EPA tested six vehicles (four conventional and
two hybrid) over the FTP, SC03, and HFET emission tests in a standard test cell at 60 F, 75 F,
and 95 F with and without the A/C system operating in order to assess the relative impact of
A/C use at city and highway speeds. The data indicated that it was more accurate to assume
that the impact of the A/C compressor on fuel consumption was the same at city and highway
speeds when compared in terms of fuel burned per unit time than when compared in terms of
fuel use per mile.  Thus, EPA estimated the impact of A/C in terms of fuel use per mile at
highway speeds by multiplying the A/C related fuel use at city speeds by the ratio of the  speed
of the city test to that of the highway test.  For average driving in the U.S., this ratio was
estimated to be 0.348.  The result was that the impact of engaging the A/C compressor 100%
of the time at highway speeds increased fuel use by 9.7%, versus 27% at city speeds. These
percentages are based on the assumptions that fuel is only consumed during warmed up
driving, hence ignoring cold start fuel use.

       EPA's estimate in the Fuel Economy Labeling rule of in-use A/C compressor
engagement was based on a test program covering 1004 trips made by 19 vehicles being
operated by their owners in Phoenix, Arizona.24 The results of this testing were correlated
against heat index, a function of temperature and humidity, and time of day, to represent solar
load.  Nationwide, EPA estimated that the A/C compressor was engaged 15.2% of the time.
However, much of this time, the ambient conditions are less severe than those of the SC03
test.  Therefore, EPA reduced this percentage to 13.3% to normalize usage to the load
experienced during SC03 conditions. On a nationwide basis, EPA estimated that the A/C
system was turned on an average of 23.9% of the time.25 Resulting in 14.3 g/mi per vehicle
CCh-equivalent impact due to A/C use (where 30% of the vehicle fleet is equipped with
automatic A/C controls, and 70% of the fleet is  equipped with manual controls).11

       This  estimate does not include defroster usage, while the NREL-OAP and ARB-
NESCCAF estimates do include this. EPA considered adding the impact of defroster usage
based in large part on NREL-OAP estimates. NREL-OAP estimates that the defroster is in-
use 5.4% of the time. However, the load of the  compressor under defrosting conditions is
very low.  EPA estimated that including defroster usage would increase the percentage of time
that the compressor was engaged at a load equivalent to that over SC03 from 13.3% to 13.7%.
While this defroster impact was quantified, EPA decided not to include it in its final 5-cycle
H Fraction of fleet equipped with automatic A/C control is based on is based on industry estimates and an EPA
analysis of the percentage of 2008 U.S. car sales - as published in the 2009 Ward's Automotive Yearbook - for
vehicle categories likely to be equipped with automatic A/C (e.g. middle luxury car, specialty, middle luxury
SUV, large luxury SUV, et. al.)
                                        2-21

-------
Draft Regulatory Impact Analysis

fuel economy formulae.  Based on the A/C usage factor of 13.3% and EPA's 5-cycle
formulae, A/C system use increases onroad fuel consumption by 2.4%. Including defroster
use modestly increased this value to 2.5%.

       Comparing the results of the three studies, the EPA estimate gives the smallest A/C
system impact, while the NREL-OAP estimate is the highest. The NESCCAF and NREL-
OAP studies give very similar results. The overall difference between the estimates is more
than a factor of two.

       It is difficult to directly compare the three estimates.  The NREL-OAP and ARB-
NESCCAF methodologies are very similar. However, the EPA methodology is quite
different, as will be discussed further below. This complicates the comparison, making it
difficult to compare smaller segments of each study directly. In addition, as will be seen, each
study utilizes assumptions or estimates which contain uncertainties. These uncertainties are
not well characterized. EPA concluded that it is not possible to determine a single best
estimate of A/C fuel use from these studies. However, EPA  was able to identify a couple of
aspects of the studies which could be improved for the purpose of this analysis. Doing so, the
overall difference between the studies was reduced by roughly one half. This process is
described below.

       The first step in this comparison will reduce the number of studies from three to two.
The NREL-OAP and ARB-NESCCAF methodologies are very similar, since both utilize the
NREL-OAP comfort model to estimate A/C usage onroad. They also both use essentially the
same estimate of A/C compressor load from Delphi to estimate the load which the compressor
puts on the engine. ARB-NESCCAF utilized the vehicle simulation tool, AVL's CRUISE
model, to estimate the impact of A/C load on fuel economy,  while NREL employed the
ADVISOR model (both models assumed a rather simple A/C system load).  In addition,
ARB-NESCCAF modeled both city and  highway driving (i.e., the 55/45 FTP/HFET), while
NREL-OAP only modeled the FTP. Thus, EPA will focus on the NESCCAF estimate over
that of NREL-OAP, though as mentioned above, their overall estimates are very similar.
Also, because NESCCAF estimated A/C fuel use nationwide, while ARE focused on that in
California, EPA will focus on comparing the NESCCAF and EPA methodologies and results
below. With respect to EPA's estimates  from the 2006 rulemaking, the estimate including
defroster use will be used, since NESCCAF considered defroster use, as well. As way of
reminder, on a nationwide average basis, the NESCAFF estimates indicate that A/C use
represents 5.3% of total fuel consumption, while EPA estimates this at 2.5%.

       NESCCAF and EPA break down the factors which determine the impact of A/C use
on onroad fuel consumption differently.  NESCCAF breaks down the process into three parts.
The first is the frequency that drivers turn on their A/C system. The second is the average
load of the A/C compressor at various ambient conditions, including compressor cycling.  The
third is the impact of this average A/C compressor load on fuel economy over various driving
conditions.

       In contrast, in the fuel labeling rulemaking, EPA breaks down the process into two
parts.  The first is the frequency that the A/C compressor is engaged at various ambient
conditions.  This includes both the frequency that the driver turns on the A/C unit and the

                                      2-22

-------
                                                                   Air Conditioning

frequency that the compressor is engaged when the system is turned on. The second is the
impact of the A/C compressor on fuel economy over various driving conditions when the
compressor is engaged.

       The most direct comparison that can be made between the two studies is the estimate
of A/C system use. Because EPA measured both A/C system on/off condition as well as
compressor engaged/disengaged condition in the Phoenix test program, it is possible to
compare the percentage of A/C system use as measured in the Phoenix study and extrapolated
to the U.S. to that of the NREL-OAP comfort model.

       In its rulemaking analysis, based on its Phoenix study and extrapolation procedure,
EPA estimated that on average, the A/C unit was turned on 23.9% of the time. This does not
include defroster use. There, EPA also determined that the NREL-OAP thermal comfort
model predicts a higher percentage of 29%, again ignoring defroster use. Since EPA utilized
NREL-OAP's estimate of defroster use in its analysis, this estimate does not contribute to the
difference in the two estimates.  Also, fuel use is very low during defroster use compared to
air conditioning at high ambient temperatures, so the difference between the 23.9% and 29%
estimates is the most relevant factor. By itself (ignoring fuel use during defrosting), this
difference would cause the NESCCAF A/C fuel use estimate to be 27% higher than that of
EPA. The overall difference between the 5.3% and 2.5% estimates is 112%.  Thus, the
difference in estimated A/C system use explains about one-fourth of the overall difference
between the two studies.

       NREL's thermal comfort model for vehicle A/C use is based on a model designed to
the represent the comfort of a person walking outside and wearing one of two different sets of
clothes. A number of assumptions had to be made in order to extrapolate this outdoor model
to a person sitting in a vehicle. The predictions of NREL-OAP's thermal comfort model have
not been confirmed with any vehicle/occupant testing and their air conditioner settings.
Therefore, its predictions, while reasonable, are of an unknown accuracy.

       EPA's Phoenix study was performed over a relatively short period of time, roughly
seven weeks. It was conducted in only one city, Phoenix. Thus, the variation in climate
evaluated was limited.  The number of vehicles tested was also fairly small, nineteen.
However, over 1000 trips were monitored by these 19 vehicles. EPA extrapolated the
measured A/C compressor engagement under these limited ambient  conditions to other
conditions using a metric called the heat index, which combines temperature and humidity
into a single metric. Heat index is conceptually similar to NREL-OAP's comfort model. This
allowed the results found in the generally dry climate of Phoenix to be extrapolated to both
cooler and more humid conditions typical of the rest of the U.S. No testing has yet been
performed to confirm the accuracy of this extrapolation.

       Given the two very different approaches to estimating vehicle A/C system use, it is
notable that the difference in the two estimates is only a relative 27%.  As both the EPA and
NREL-OAP models of A/C system use involve assumptions or extrapolations which have not
been verified, it is not possible to determine which one is more accurate. Thus, the
differences in the EPA and ARE estimates of the impact of A/C use  on onroad fuel
consumption due to these two different sources of A/C usage cannot be resolved at this time.

                                        2-23

-------
Draft Regulatory Impact Analysis

       With respect to the operation of the A/C compressor at various ambient and driving
conditions, EPA bases its estimate on the Phoenix vehicle test study.  This is subject to the
same uncertainties described above, due mainly to the limited scope of the data. NREL-OAP
relies on test results published by W.O. Forrest of Delphi. Forrest describes the factors which
affect the load of the A/C system on the engine: the percentage of time the compressor is
engaged, compressor displacement, compressor speed, air flow across the evaporator, engine
operating condition and ambient conditions.  The load curves presented by Forrest apply to a
210 cc compressor and show load as a function of compressor speed for six sets of ambient
conditions. The loads include the effect of compressor cycling.  However, no mention is
made of airflow rates across the evaporator, which would vary with engine speed.  It is not
clear whether these curves were based on bench testing or onroad vehicle testing.  Also, only
one A/C system appears to have been tested. It is not clear how well these curves would
apply to other manufacturers' systems, nor even to others produced by Delphi. Forrest states
that the loads for other compressor displacements can be approximated by assuming that the
load is proportional to compressor displacement.  However, this is clearly an approximation
and does not address differences inherent in particular A/C  system applications. The fact that
the NESCCAF analysis is based on the testing of only a single A/C system and does not
address the effect of varying airflow rates under different driving conditions appears to be the
largest sources of uncertainty in their estimate.

       It is not possible to directly compare these two estimates of compressor operation.
EPA's Phoenix study provides an estimate of the  percentage of time that the compressor is
engaged when the A/C system  is on. On the other hand, compressor cycling is implicitly
included in the Delphi  load curves. Since the load curves of a continuous operating
compressor were not presented, the degree of cycling cannot be determined. Thus, the effect
of any  differences in the NESCCAF and EPA estimates of compressor engagement cannot be
quantified.

       With respect to the impact of the A/C compressor load on fuel economy, EPA relies
on a comparison of measured fuel economy over  the two warmed up bags of its FTP test
(when the  A/C system is inoperative) and its SC03,  A/C emissions test.  The vehicles  on both
tests are run at city speeds.  EPA based its estimates on the testing of over 600 recent model
year vehicles. Thus, for the conditions addressed by the SC03 test, EPA's estimate of the
impact of A/C system load on fuel economy  is well supported. However, in order to combine
this measurement with the Phoenix study, EPA needed an estimate of the percentage of time
that the compressor was engaged during the SC03 test. The SC03 test does not include a
measurement of this factor, so EPA had to estimate  the percentage of  time that the compressor
was engaged during the test. As noted above, EPA  assumed that the A/C compressor was
engaged 100% of the time during the SC03 test given its  short duration and the pre-heating of
the vehicle. Thus, for a given ambient condition, if the compressor was  estimated to be
engaged 25% of the time, then the incremental amount of fuel used due to A/C system was
25% of the difference between the fuel use over the SC03 test and a 39%/61% weighting of
the fuel use over Bags 2 and 3 of the FTP, respectively.

       EPA has evidence to show that most vehicles' A/C compressors  are engaged 100% of
the time over SC03.26  The vehicle pre-heating, short test duration and the requirement that
the driver window be rolled down, make it extremely likely that the vehicle compartment

                                        2-24

-------
                                                                   Air Conditioning

never reaches a comfortable temperature by the end of the test. However, it is possible that
the compressor still cycled to some degree during the test. All compressors shut down when
the heat exchanger nears 32 F in order to avoid icing. The cold heat exchanger continues to
cool the refrigerant while the compressor is shut down, but the compressor is not putting an
additional load on the engine and increasing fuel  consumption. As it is impossible for the
compressor to operate more than 100% of the time, any error in EPA's assumption can only
lower the actual compressor use below 100%. If compressor engagement was lower than
100%, this would mean that fuel use at 100% compressor engagement would be higher than
currently estimated.  Thus, it is possible that this assumption that the A/C compressor is
engaged 100% during SC03 is causing EPA's estimate of A/C fuel use to be under-estimated
to some degree.

       There are additional uncertainties involved in EPA's assumption that a vehicle's A/C
fuel use is constant in terms of gallons per hour, and thus inversely proportional to vehicle
speed when presented in terms of gallons per mile. EPA testing of six vehicles as part of the
Fuel Economy Labeling rulemaking (used to estimate A/C compressor usage in highway
driving conditions, as noted above) confirmed that A/C fuel use was roughly constant in terms
of gallons per hour. However, this testing was performed in a standard emission test cell.  Air
flow through the engine compartment was the same at city and highway speeds. The city test
was only 20 minutes long and the highway test was only  10 minutes long. There was also
significant variability in the individual vehicle test results. Thus, while the testing showed
that EPA's assumption was reasonable, there is an unknown degree of uncertainty associated
with extrapolating the measured A/C fuel use at city  speeds to highway speeds.  One could
attempt to quantify the uncertainty using the test results of the six vehicles.  However, these
vehicles were not randomly selected and two of the six vehicles were Prius hybrids.  Thus, it
is not clear how representative the results of a statistical analysis of these data would be.

       An A/C load adjustment factor is also applied to account for the change in compressor
load which occurs when the compressor is engaged at different temperatures.  The study
which developed this data data is based on an A/C model developed by Nam (2000).27

       NESCCAF starts with A/C compressor load curves which describe the A/C
compressor load as a function of compressor speed for six ambient conditions. These curves,
along with A/C - on percentages from the thermal comfort model,  were used to interpolate
between the six compressor load curves to estimate the load curves applicable to the ambient
conditions existing during driving times for a large number of cities across the U.S. The
resulting curves are averaged using the VMT estimated to occur in each city to produce a
single load curve representing the entire U.S.

       NESCCAF then input this national average load curve into AVL's CRUISE model to
estimate the effect of A/C on fuel consumption over the FTP and HFET cycles.  The CRUISE
model simulates vehicle operation and fuel consumption over specified driving conditions.
The load of the A/C compressor (based on bench testing) was added to the other loads being
placed on the vehicle, such as inertia, friction, aerodynamic drag, etc. The A/C loads included
the cycling of the compressor as a function of ambient condition. In actuality, the engine will
experience the full load of the compressor at some times and no load at other times. This
could produce a slightly different fuel use impact than applying the average load of the

                                       2-25

-------
Draft Regulatory Impact Analysis

compressor all of the time.  However, this error is likely very small. The A/C load curves
vary as a function of engine speed, but not vehicle speed.  However, as air flow by the heat
exchanger will vary as a function of vehicle speed, compressor cycling and evaporator
cooling efficiency is likely to vary, as well. However, the degree of error associated with any
of these simplifications is unknown.

       A detailed comparison of this aspect of the two analyses would require reconstructing
both models to produce A/C fuel use estimates for specific ambient conditions. This is
beyond the scope of the study.  Also, once the differences were known, it would still be
difficult to decide which estimate was superior.

       There is one aspect of each analysis which appears to be an improvement over the
other. In addition to A/C, EPA evaluated a number of other reasons why onroad fuel
economy differs from that measured over the FTP and HFET cycles. Among these were
higher speed and more aggressive driving, ambient temperatures below 75 F, short trips,
wind, under-inflated tires, ethanol containing fuel, etc. This does not affect the absolute
volume of fuel used by the A/C system, but it does raise the total amount of fuel consumed
onroad, effectively lowering the percentage of fuel due to A/C use.

       NESCCAF estimated the impact of the A/C compressor load on fuel use during city
and highway driving using the CRUISE model. While it is not clear that this is superior to
EPA's SC03 data, the CRUISE model is likely more accurate for highway driving than an
extrapolation of the SC03 data (i.e. EPA's six vehicle study  described above).  While
CRUISE was not able to represent all aspects of vehicle operation, such as airflow across the
evaporator, it does simulate the difference in engine speed and load between city and highway
driving. This allows a detailed simulation of the A/C compressor speed during this driving,
which is a primary factor in estimating A/C compressor load. EPA's extrapolation of the
impact over SC03 essentially assumes that engine speed and airflow over the evaporator are
the same during both city and highway driving, or that any differences cancel each other.
This is unlikely.  Therefore, NESCCAF's highway estimates are likely more accurate than
EPA's.

       Since the two analyses were performed so differently, the CRUISE results for highway
driving cannot be simply substituted for EPA's estimates. However, one way to utilize the
CRUISE highway results is to determine the ratio of the impact of the A/C load on fuel use
over the HFET to that over the FTP. This ratio can then be substituted for EPA's  assumption
that the impact of A/C load is constant with time (inversely proportional to vehicle speed in
terms of gallons per mile.

       Adjusting the NESCCAF estimates for the other factors reducing onroad fuel economy
relative to the FTP/HFET is straightforward.  EPA found that all such factors, including A/C,
reduced onroad fuel economy to 80% of the FTP/HFET.  In other words, onroad fuel
consumption is 25% higher (1/0.8) than over the FTP/HFET. Thus, the CO2 emissions over
the FTP/HFET shown above in Table 2-7 are multiplied by a factor of 1.25 to represent
onroad CCh emissions. A/C fuel use is unaffected. A/C fuel use as a percentage of onroad
fuel use is simply the ratio of the A/C fuel use divided by the estimated onroad fuel use.
                                        2-26

-------
                                                                  Air Conditioning

These figures are shown in Table 2-9 below. The VMT weighted average of these
percentages is 4.4%, 0.9% lower than the estimate presented above.

       Table 2-9: Adjusted NESCCAF CO2 Emissions Over 55/45 FTP/HFET Tests and From A/C Use
                                       (g/mi)

55/45 FTP/HFET
Indirect A/C Fuel Use
Indirect A/C Fuel Use
Small Car
349
16.8
4.8%
Large Car
413
19.1
4.6%
Minivan
472
23.5
5.0%
Small Truck
535
23.5
4.4%
Large Truck
619
23.5
3.8%
       Incorporating the relative impact of A/C load on fuel consumed over the HFET versus
FTP cycles from CRUISE requires a few steps. Table 2-10 shows the incremental CCh
emissions from the A/C compressor load from the CRUISE simulations of the FTP and HFET
cycles. The top half of the table shows the incremental fuel use in terms of grams CCh per
mile.  These figures were taken from Tables B-20 through B-23 of the NESCCAF report.28
For the large car, two base vehicles were simulated. EPA selected the vehicle with the
conventional gasoline engine with variable valve  timing and lift. The large truck was not
modeled using CRUISE. Further in the study, Meszler assumed that the A/C fuel impact was
proportional to compressor displacement. The large truck is assumed to have the same
compressor displacement as the minivan and small truck.  Thus, the A/C fuel impact was
estimated for the large truck as the average of the impacts for the minivan and small truck.
The bottom half of the table shows the incremental fuel use in terms of grams  CCh per
minute. These figures were calculated by multiplying the A/C fuel impacts in grams per mile
by the average speeds of the FTP and HFET cycles: 19.6 and 48.2 mph and converting hours
to minutes. The final line of the table  shows the ratio of the incremental fuel use in terms of
grams CCh per minute for the HFET cycle to that over the FTP.

                          Table 2-10: Impact of A/C on Fuel Use: System

Small Car
Large Car
Minivan
Small Truck
Large Truck
A/C impact: 100% A/C System On Time (g/mi)
FTP
HFET
67.4
32.3
56.6
31.9
81.8
45.0
89.7
47.4
85.8
46.2
A/C impact: 100% A/C System On Time (g/minute (g/min))
FTP
HFET
HFET/FTP (g/min)/(g/min)
22.02
25.95
1.18
18.49
25.63
1.39
26.7
36.2
1.35
29.3
38.1
1.30
28.0
37.1
1.32
       As can be seen in the last line of Table 2-10, the ratio of A/C CO2 emissions over the
HFET to that over the FTP is greater than 1.0 for each of the five vehicles. VMT weighting
the CO2 emissions for each of the five vehicle groups produces an average ratio of 1.30.  EPA
assumed that this ratio was 1.0.  Thus, EPA likely underestimated the impact of A/C fuel use
during highway driving by 30%. For the purposes of EPA's onroad fuel economy labeling
rule, this under-estimation is small, because the impact of A/C on highway fuel economy is
small. However, when estimating the impact of A/C fuel use, the difference is more
significant. EPA's five cycle formulae for estimating onroad fuel economy was adjusted to
reflect this 1.32 factor. The impact of A/C fuel use on onroad fuel economy including
defrosting increased from 2.5% to  2.8%. Thus, instead of a range of 2.5-5.3% for the impact
                                       2-27

-------
Draft Regulatory Impact Analysis

of A/C on onroad fuel consumption, the range is now 2.8-4.4%. The difference between the
two estimates has been cut almost in half.

       There is one more adjustment that should be made to both estimates. Both EPA and
NESCCAF assume that all A/C systems are in working condition. However, A/C systems do
leak refrigerant, sometimes to the point where the system no longer works. Since the cost of
repairing a leak can be significant, some vehicle owners do not always choose to repair the
system. For its MOBILE6 emission model, EPA estimated the percentage of vehicles on the
road with inoperative A/C systems as a function of vehicle age. Coupling these estimates
with the amount of VMT typically driven by vehicles as a function of age, EPA estimates that
8% of all the VMT in the U.S. is by vehicles with inoperative A/C systems.  These systems do
not impact fuel consumption. Thus, both the NESCCAF and EPA estimates should be
multiplied by 0.92. Doing this, the impact of A/C on onroad fuel consumption is estimated to
be2.6-to-4.1%.

2.3.2  Technologies That  Improve Efficiency of Air Conditioning and Their
      Effectiveness

       EPA estimates that the CCh emissions from A/C related load on the engine accounts
for about 3.9% of total greenhouse gas emissions from passenger vehicles in the United
States. This is equivalent to  CCh emissions of approximately 14 g/mi per vehicle. The A/C
usage is inherently higher in hotter months and states; however, vehicle owners may use the
A/C systems throughout the year in all parts of the nation. That is, people use A/C systems to
cool and dry the cabin air for passenger comfort on hot humid days, as well as to de-humidify
the air used for defogging/de-icing the  front windshield to improve visibility.

       Most of the excess load on the engine comes from the compressor, which pumps the
refrigerant around the system loop.  Significant additional load on the engine may also come
from electrical or hydraulic fan units used for heat exchange across the condenser and
radiator.  The controls that EPA believes manufacturers would use to earn credits for
improved A/C efficiency would focus primarily, but not exclusively, on the compressor,
electric motor controls, and system controls which reduce load on the A/C system (e.g.
reduced 'reheat' of the cooled air and increased of use  recirculated cabin air). EPA is
proposing a program that would result  in improved efficiency of the A/C system (without
sacrificing passenger comfort) while improving the fuel efficiency of the vehicle, which has a
direct impact on CCh emissions.

       The cooperative IMAC program described above has demonstrated that average A/C
efficiency can be improved by 36.4% (compared to a baseline A/C system), when utilizing
"best-of-best" technologies.  EPA considers a baseline A/C system contains the following
components and technologies; internally-controlled fixed displacement compressor (in which
the compressor clutch is controlled based on 'internal'  system parameters, such as head
pressure, suction pressure, and/or evaporator outlet temperature); blower and fan motor
controls which create waste heat (energy) when running at lower speeds;  thermostatic
expansion valves; standard efficiency evaporators and  condensers; and systems which
circulate compressor oil throughout the A/C system. These baseline systems are also
extraordinarily wasteful in their energy consumption because they add heat to the cooled air

                                        2-28

-------
                                                                   Air Conditioning

out of the evaporator in order to control the temperature inside the passenger compartment.
Moreover, many systems default to a fresh air setting, which brings hot outside air into the
cabin, rather than recirculating the already-cooled air within the cabin.

       The IMAC program indicates that improvements can be accomplished by a number of
methods related only to the A/C system components and their controls including: improved
component efficiency, improved refrigerant cycle controls, and reduced reheat of the cooled
air. The program EPA is proposing would encourage the reduction of A/C CCh emissions
from cars and trucks by 40% from current baseline levels through a credit system. EPA
believes that the component efficiency improvements demonstrated in the IMAC program,
combined with improvements in the control of the supporting mechanical and electrical
devices (i.e. engine speeds and electrical heat exchanger fans), can go beyond the IMAC
levels and achieve a total efficiency improvement of 40%.  The following sections describe
the technologies EPA believes manufacturers can use to attain these efficiency improvements.

   2.3.2.1    Reduced Reheat Using a Externally-Controlled, Variable-Displacement
            Compressor

       The term 'external control' of a variable-displacement compressor is defined as a
mechanism or control strategy  where the displacement of the compressor adjusted
electronically, based on the temperature setpoint and/or cooling demand of the A/C system
control settings inside the passenger compartment. External controls  differ from 'internal
controls' that internal controls adjust the displacement of the compressor based on conditions
within the A/C system, such has head pressure, suction pressure, or evaporator outlet
temperature. By controlling the displacement of the compressor by external means, the
compressor load can be matched to the cooling demand of the cabin.  With internal controls,
the amount of cooling delivered by the system may be greater than desired, at which point the
cooled cabin air is then 'reheated' to achieve the desired cabin comfort.  It is this reheating of
the air which results reduces the efficiency of the A/C system - compressor power is
consumed to cool air to a temperature less than what is desired.

       Reducing reheat through external control of the compressor is a very effective strategy
for improving A/C system efficiency. The SAE IMAC team determined that an annual
efficiency improvement of 24.1% was possible using this technology.29 EPA estimates that
additional improvements with this technology, when fully developed, calibrated, and
optimized to particular vehicle's cooling needs - and combined with increased use of
recirculated cabin air - can result in an efficiency improvement of 40%, compared to the
baseline system.

    2.3.2.2  Reduced Reheat Using a Externally-Controlled, Fixed-Displacement or
            Pneumatic Variable-Displacement Compressor

       When using a fixed-displacement or pneumatic variable-displacement compressor
(which controls the stroke, or displacement, of the compressor based on system suction
pressure), reduced reheat can be realized by disengaging the compressor clutch momentarily
to achieve the desired evaporator air temperature. This disengaging, or cycling, of the
compressor clutch must be externally-controlled in a manner similar to that described in

                                        2-29

-------
Draft Regulatory Impact Analysis

2.3.2.1. EPA believes that a reduced reheat strategy for fixed-displacement and pneumatic
variable-displacement compressors can result in an efficiency improvement of 20%. This
lower efficiency improvement estimate (compared to an externally-controlled variable
displacement compressor) is due to the thermal and kinetic energy losses resulting from
cycling a compressor clutch off-and-on repeatedly.

    2.3.2.3   Defaulting to Recirculated Cabin Air

       In ambient conditions where air temperature outside the vehicle is much higher that
the air inside the passenger compartment, most A/C systems draw air from outside the vehicle
and cool it to the desired comfort level inside the vehicle.  This approach wastes energy
because the system is continuously cooling the hotter outside air instead of having the A/C
system draw it's supply air from the cooler air inside the vehicle (also known as 'recirc'). By
only cooling this inside air (i.e. air that has been previously cooled by the A/C system), less
energy is required, and A/C Idle Tests conducted by EPA indicate that an efficiency
improvement 30% improvement is possible.  A mechanically-controlled door on the A/C
system's air intake typically controls whether outside air, inside air, or a mixture of both, is
drawn into the system.  Since the typical 'default' position of this air intake door is outside air
(except in cases where maximum cooling capacity is required, in which case, many systems
automatically switch this door to the recirculated air position), EPA is proposing that, as cabin
comfort and de-fogging conditions allow, an efficiency credit be granted if a manufacturer
defaults to recirculated air whenever the outside ambient temperature is greater than 75°F. To
maintain the desired quality inside the cabin  (in terms of freshness and humidity), EPA
believes some manufacturers will equip their A/C systems with humidity sensors, which will
allow them to adjust the blend of fresh-to-recirculated air and optimize  the controls for
maximum efficiency.

    2.3.2.4   Improved Blower and Fan Motor Controls

       In controlling the speed of the direct current (DC) electric motors in an air
conditioning system, manufacturers often utilize resistive elements to reduce the voltage
supplied to  the motor, which in turn reduces  its speed.  In reducing the voltage however, these
resistive elements produce heat, which is typically dissipated into the air ducts of the A/C
system. Not only  does this waste heat consume electrical energy, it contributes to the heat
load on the A/C system. One method for controlling DC voltage is to use a pulsewidth
modulated (PWM) controllers for each motor. A PWM controller can reduce the amount of
energy wasted, and based on Delphi estimates of power consumption for these devices, EPA
believes that when more efficient speed controls are applied to both the blower and fan
motors, an overall improvement in A/C system efficiency of 15% is possible.30

    2.3.2.5   Electronic Expansion Valve

       The expansion valve in an A/C system is used to "throttle" the flow high pressure
liquid refrigerant upstream of the evaporator. By throttling the refrigerant flow, it is possible
to control the amount of expansion (superheat) that the refrigerant will undergo, and by
extension, the amount of heat removed from air passing through the evaporator. With a
conventional, or thermostatic, expansion valve (TXV), the amount of expansion is controlled

                                         2-30

-------
                                                                     Air Conditioning

by an internal temperature reference to assure a constant temperature level for the expanded
refrigerant gas, which is typically a few degrees Celsius above the freezing point of water
(which may be too cool for the desired cabin comfort level). In the case where the air exiting
the evaporator is too cool (or over-cooled), it will be necessary to reheat it by directing some
of the airflow through the heater core.  It is this reheating of the air which results in reduced
system efficiency, as additional compressor energy is consumed in the process of over-
cooling the air.  However, if the expansion of the refrigerant is controlled externally - such as
by an electronic signal from the A/C control unit - it is possible to adjust the level of
expansion, or superheat, to only to the level necessary to meet the current cooling needs of the
passenger compartment. This electronic expansion valve (EXV) approach is similar to the
reduced reheat strategy, except that instead of controlling the mass  of refrigerant flowing
through the system by controlling the compressor output, the mass flow is controlled by the
EXV. By reducing the amount of refrigerant expanding, or controlling the level of superheat
in the gas-phase refrigerant, the temperature of the evaporator can be increased and controlled
to the point where reheating of the air is not necessary, the SAEIMAC team determined that
an annual efficiency improvement of 16.5% is possible. EPA estimates the when fully
developed, calibrated, and optimized to the requirements of particular system design, use of
EXV technology can result in a 20% efficiency improvement over the baseline TXV system.

    2.3.2.6  Improved-Efficiency Evaporators and Condensers

       The evaporators and condensers in an  A/C system are designed to transfer heat to and
from the refrigerant - the  evaporator absorbs heat from the cabin air and transfers it to the
refrigerant, and the condenser transfer heat from the refrigerant to the outside ambient air.
The efficiency, or effectiveness, of this heat transfer process directly  effects the efficiency of
the overall system, as more work, or energy, is required if the process is inefficient. A
method for measuring the heat transfer effectiveness of these components is to determine the
Coefficient of Performance (COP) for the system using the industry-consensus method
described in the SAE surface vehicle standard J2765 - Procedure for Measuring System COP
of a Mobile Air Conditioning System on a Test Bench.31 If these components can
demonstrate a 10% improvement in COP versus the baseline components, EPA estimates that
a 20% improvement in overall system efficiency is possible.

    2.3.2.7  Oil Separator

       The oil present in a typical A/C system circulates throughout the system for the
purpose of lubricating the compressor.  Because this oil is in contact with inner surfaces of
evaporator and condenser, and a coating of oil reduces the heat transfer effectiveness of these
devices, the overall system efficiency is reduced.32 It also adds inefficiency to the system to
be "pushing around and cooling" an extraneous fluid that results in a dilution of the
thermodynamic properties of the refrigerant.  If the oil can be contained only to that part of
the system where it is needed - the compressor - the heat transfer effectiveness of the
evaporator and condenser will improve. The overall COP will also improve due to a
reduction in the flow of diluent. The SAE IMAC team estimated that overall system COP
could be improved by 8% if an oil separator was used.14 EPA believes that if oil is prevented
from prevented from circulating throughout the A/C system, an overall system efficiency
improvement of 10% can be realized.

                                        2-31

-------
Draft Regulatory Impact Analysis

2.3.3 Technical Feasibility of Efficiency-Improving Technologies

       EPA believes that the efficiency-improving technologies discussed in the previous
sections are available to manufacturers today, are relatively low in cost, and that their
feasibility and effectiveness has been demonstrated by the SAEIMAC teams and various
industry sources.  EPA also believes that when these individual components and technologies
are fully designed, developed, and integrated into A/C system designs, manufacturers will be
able to achieve the estimated reductions in CC>2 emissions and earn appropriate A/C
Efficiency Credits, which are discussed in the following section.

2.3.4 A/C Efficiency Credits

       In model years 2012 through and 2016, manufacturers would be required to
demonstrate that vehicles receiving credit for A/C efficiency improvements are equipped with
the type of components and/or controls needed to qualify for a certain level of CC>2 credit.
For model years 2014 and later, the design-based approach will be supplemented with a
vehicle performance test, which has been modified slightly from that proposed in the GHG
Mandatory Reporting Rule. In particular, EPA is proposing that the range of allowable
ambient temperature for a valid A/C Idle Test be limited to 75 + 2 °F  (as opposed to 68-to-86
°F for a valid FTP test) and that the humidity in the test cell be limited to 50 + 5 grains of
water per pound of dry air (where there are no such humidity constraints on an FTP test, only
a humidity correction for NOx).  This narrowing of the allowable range of ambient conditions
was done to improve the accuracy and repeatability of the test results.  Since the performance
of an A/C system  (and the amount of fuel consumed by the A/C system) are directly
influenced by the  heat energy, or  enthalpy, of the air within the  test cell - where criteria
pollutants are not  - it was necessary to control the enthalpy, and limit its effect on the test
results.  In addition, EPA is proposing a modification to the interior fan settings  for vehicles
with manual A/C controls.  In the proposed reporting  rule, vehicle with manual A/C controls
were to be run on  the 'high' fan setting for the duration of the A/C on portion of the test.
However, EPA believes that this fan speed setting would unduly penalize vehicles with
manual controls when compared to those with automatic control - as automatic controls adjust
the fan speed to lower setting as the target interior temperature is reached (which is similar to
what a driver does on a vehicle with manual controls). In recognition of this disparity in the
proposed test procedure, EPA is revising the test to allow vehicles with manual A/C controls
to average the result obtained on the high fan speed setting with the result obtained on the low
fan speed setting.  The additional 10-minute idle sequence on the low fan speed  setting is to
be run immediately following the high fan sequence (no additional prep cycle is required).
This revised performance test will assure that the A/C components and/or system control
strategies a manufacturer chooses to implement are indeed delivering the efficiency gains
projected for each. The performance test discussed in section II of the preamble is the A/C
Idle Test, but in that section, EPA also discusses how a modified SC03 test could also be used
to measure the efficiency of A/C systems.

       To establish an average A/C CO2 rate for the A/C systems in todays vehicles, the EPA
conducted laboratory tests to measure the amount of additional  CCh a vehicle generated due
to A/C use on the  proposed Idle Test.33  The results of this test program are summarized in
Table 2-11, and represent a wide cross-section of vehicle types  in the U.S. market.  The

                                        2-32

-------
                                                                          Air Conditioning

average A/C €62 result from this group of vehicles is the value against which results from
vehicle testing (beginning in 2014) will be compared. The EPA conducted laboratory tests to
tested over 60 vehicles representing a wide range of vehicle types (e.g. compact cars, midsize
cars, large cars, sport utility vehicles, small station wagons, and standard pickup trucks).

    Table 2-11 Summary of A/C Idle Test Study Conducted by EPA at the National Vehicle Fuel and
                                    Emissions Laboratory

Vehicle Makes Tested	19	
Vehicle Models Tested	29	
Model Years Represented (number of vehicles in each model   1999 (2), 2006 (21), 2007 (39)
year)	
EPA Size Classes Represented                           Minicompact, Compact, Midsize, and
                                                    Large Cars
                                                    Sport Utility Vehicles
                                                    Small Station Wagons
                                                    Standard Pickup Trucks
Total Number of A/C Idle Tests	62	
Average A/C CO2 (g/min)	21.3	
Standard Deviation of Test Results (+ g/min)                5.8
                                            2-33

-------
Draft Regulatory Impact Analysis

       The majority of vehicles tested were from the 2006 and 2007 model years and their
A/C systems are representative of the 'baseline' technologies, in terms of efficiency (i.e. to
EPA's knowledge, these vehicles do not utilize any of the efficiency-improving technologies
described in Table 2-12). The individual test results from this testing are shown in Figure 2-4.
EPA attempted to find a correlation between the A/C CO2 results and a vehicle's interior
volume, footprint, and engine displacement, but was unable to do so, as there is significant
"scatter" in the test results.  This scatter is generally not test-to-test variation, but scatter
amongst the various vehicle models and types - there is no clear correlation between which
vehicles perform well on this test, and those which do not. EPA did attempt to find a
correlation between the idle test results and a vehicle's interior volume, footprint, or engine
displacement, but no clear correlation could be found.  What is clear, however, is that load
placed on the engine by the A/C system is not consistent, and in certain cases, larger vehicles
perform better than smaller ones, in terms of their A/C CCh result.
                    Average A/C CO2 rate = 21.3 g/min
               •••«
                                            30         40
                                               Test Number
              Figure 2-4 EPA A/C Idle Test Results from Various Vehicle Model Types

       Part of this variation in the proposed A/C Idle Test results may be due to the
components a manufacture chooses to use in a particular vehicle. Where components such as
compressors are shared across vehicle model types (e.g. a compressor may be 'over-sized' for
one application, but the use of a common part amongst multiple model types results in a cost
savings to the manufacturer).  Some of the variation may also be due to the amount of cooling
capacity a vehicle has at idle. One manufacturer indicated that one of their vehicles which
produced a below-average A/C CCh result, is also known for having A/C performance at idle
which does not meet customer expectations, but off-idle, performs very well. Therefore, it
                                         2-34

-------
                                                                    Air Conditioning

will be necessary for manufacturers to balance the cooling capacity of the A/C system under
idle conditions against the overall A/C system efficiency.

       Some of this variation between various models may also be due to the efficiency of the
fan(s) which draw air across the condenser - since an external fan is not placed in front of the
vehicle during the A/C Idle Test, it is the vehicle's fan which is responsible for rejecting heat
from the condenser (and some models may do this more efficiently than others). In this case,
EPA believes that an SC03-type test - run in a full environmental chamber with a "road-
speed" fan on the front of the vehicle - would be a better measure of how a vehicle's A/C
system performs under transient conditions, and any limitations the system may have at idle
could be counter-balanced by improved performance and efficiency elsewhere in the drive
cycle. However, since idle is significant part of real-world and FTP drive cycles (idle
represents 18% of the FTP), EPA believes that the focus in this rulemaking on A/C system
efficiency under idle conditions is justified.

       The average A/C CCh result for the vehicles tested was 21.3 g/min.  Starting in the
year 2014, in order to qualify for A/C Efficiency Credits, it will be necessary for
manufacturers to demonstrate the efficiency of their systems by running an A/C Idle Test on
each vehicle model for which they are seeking credit. To qualify for credit, it will be
necessary for each model to achieve an A/C CCh result less than or equal to  14.9 g/min
(which is 30% less than the average value observed in the EPA testing). EPA chose the 30%
improvement over the "average" value to drive the fleet of vehicles toward A/C systems
which approach or exceed the efficiency of current best-in-class vehicles. EPA believes this
approach will cause  manufacturers to tailor the size A/C components and systems to the
cooling needs of a particular vehicle model and focus on the overall efficiency of their A/C
systems.  EPA believes this approach strikes a reasonable balance between avoiding granting
credits for improvements which would occur in any case, and encouraging A/C efficiency
improvements which would not otherwise occur. Once manufacturers begin using the
technologies described in Table 2-12 - and develop these technologies for the requirements of
each vehicle, with a focus on achieving optimum efficiency - EPA believes it will be possible
to demonstrate that a vehicle is indeed achieving the reductions in A/C CCh emissions that are
estimated for this rulemaking.

       We believe that it is possible to identify the A/C efficiency-improving components
and control strategies most-likely to be utilized by manufacturers and are assigning a CCh
'credit' to each. In addition, EPA recognizes that to achieve the maximum efficiency benefit,
some  components can be used in conjunction with other components or control strategies.
Therefore, the system efficiency synergies resulting from the grouping of three or more
individual components are additive, and will qualify for a credit commensurate with their
overall effect on A/C efficiency. A  list of these technologies - and the credit associated with
each - is shown in Table 2-12.  If the more than one technology is utilized by a manufacturer
for a given vehicle model, the A/C credits can be added, but the maximum credit possible is
limited to 5.7 g/mi.  This maximum credit represents a 40% improvement over a 14.3 g/mi
per vehicle CCh-equivalent impact due to A/C use.  This 14.3 g/mi impact is derived from the
EPA's 2006 estimate of fuel consumption due to A/C use of 12.11 g/mi.  However, the 2006
estimate needed to be adjusted upward to reflect the  increased prevalence of "automatic" A/C
controls in modern vehicles  (the Phoenix study used in the EPA's 2006 estimate was from

                                         2-35

-------
Draft Regulatory Impact Analysis

1990s-vintage vehicles, which do not include a significant number of vehicles with automatic
climate control systems).  To derive the newer estimate, a scenario was first modeled in which
100% of vehicles used in the Phoenix study were equipped with automatic A/C systems
(which increases the amount of time the compressor is engaged in moderate ambient
conditions), which resulted in the 12.11 g/mi estimate increasing to 17.85 g/mi. Industry and
supplier estimates were then used for the number of vehicles equipped with automatic A/C
systems - as well as vehicle sales data from the 2009 Ward's Automotive Yearbook - and
projected that 38% of new vehicles are equipped with automatic A/C systems.34 Finally, the
percentages of vehicles with and without automatic A/C systems were multiplied by their
respective impact on fuel consumption (0.62 x 12.11 + 0.38 x 17.85) to produce our estimate
of 14.3 g/mi. This credit is the same for cars and trucks because the A/C components, cooling
requirements, and system functions are similar for both vehicle classes.  Therefore, EPA
believes the level of efficiency improvement and the maximum credit possible should be
similar for cars and trucks as well.

                Table 2-12: Efficiency-Improving A/C Technologies and Credits
Technology Description
Reduced reheat, with externally -controlled,
variable-displacement compressor
Reduced reheat, with externally -controlled,
fixed-displacement or pneumatic variable
displacement compressor
Default to recirculated air whenever ambient
temperature is greater than 75 °F
Blower motor and cooling fan controls which
limit waste energy (e.g. pulsewidth modulated
power controller)
Electronic expansion valve
Improved evaporators and condensers (with
system analysis on each component indicating
a COP improvement greater than 10%, when
compared to previous design)
Oil Separator
Estimated Reduction
in A/C CO2 Emissions
30%
20%
30%
15%
20%
20%
10%
A/C Credit (g/mi
CO2)
1.7
1.1
1.7
0.9
1.1
1.1
0.6
       The estimates for the percent reduction in A/C CO2 for each technology is based in
part on the results of S AE IMAC Team 2 (Improved Efficiency) final report, which both
provides a baseline for calculating creditable improvements, and also provides a level of
improvement for each technology.  The estimated percent reduction in A/C CO2 emissions for
                                        2-36

-------
                                                                    Air Conditioning

each was adjusted upward to reflect continuous improvement in the design, calibration, and
implementation of these technologies. These technologies, which, when combined, can allow
manufacturers to achieve the 40% reduction in CCh emissions.
2.4 Costs of A/C reducing technologies

       This section describes the cost estimates for reductions in air conditioner related GHG
emissions as well as the cost savings that result from improved technologies. These estimates
are largely determined from literature reviews of publications and public presentations made
by parties involved in the development and manufacture of A/C systems as well as from EPA
analyses. The cost savings are estimated from the literature as well as the supplemental
deterioration models based analysis described above.

       For leakage, or direct, emissions, EPA assumes that reductions can be achieved
without a change in refrigerant, though it is possible that by 2020 a new technology and
refrigerant will be a much more viable option than it is today.  For example, an alternative
refrigerant with a GWP less than 150 and can be used directly in current A/C systems will be
able to meet the leakage credit requirements without significant engineering changes or cost
increases. However,  in order to reduce the leakage in conventional R134a systems by 50%, it
has been estimated that the manufacturer cost would increase by $ 15 per vehicle in 2002
dollars, employing existing off-the-shelf technologies such as the ones included in the J2727
leakage charts.1 Converting this to 2007 dollars using the GDP price deflator (see Appendix
3 .A of the Draft Joint TSD) results in a cost of $ 17. With the indirect cost markup factor of
1.11 for a low complexity technology the compliance cost becomes $19. Using this as the
2012MY cost and applying time based learning results in a 2016MY cost of $17 for leakage
reduction technology. Table 2-13 shows how these costs may be distributed on a year by year
basis as the proposed program phases in over 5 years.

       We  expect that a reduction in leakage will  lead to fewer servicing events for
refrigerant recharge.  In 2006, the EPA estimated the average cost to the vehicle owner for a
recharge maintenance visit was $100. However, recent information indicates that the industry
average cost of recharging an automotive air conditioner is $147.35 With the new AC
systems, such $100 or $147 maintenance charges could be moved delayed until later in the
vehicle life and, possibly, one of more events could be eliminated completely.  This provides
potential savings to consumers as a result of the new technology. Note that these potential
maintenance savings  are not included in the cost and benefit analysis presented in Chapters 6
and 8 of this DRIA. However, EPA intends to include an estimate of maintenance savings in
the final rule analysis and believe that this higher estimate for the cost of recharging an A/C
system would serve as the basis for those maintenance savings in the cost analysis of the final
rule.
1 Author unknown, Alternative Refrigerant Assessment Workshop, SAE Automotive Alternative Refrigerant
Symposium, Arizona, 2003.
                                        2-37

-------
Draft Regulatory Impact Analysis
       For indirect CCh emissions due to A/C, it has been estimated that a 25-30% reduction
can be achieved at a manufacturer cost of 44€, or $51 in 2005 dollars/ The IMAC Efficiency
Improvement team of the Society of Automotive Engineers realized an efficiency
improvement of 36.4% based on existing technologies and processes.29  For the idle test, EPA
estimates that further reductions with software controls can achieve a total reduction of 40%.
Converting the $51 value to 2007 dollars results in $54 (using the GDP price deflator as
explained in Appendix 3. A of the Draft Joint TSD) and applying a 1.11 indirect cost
multiplier for a low complexity technology (as described in Chapter 3 of the Draft Joint TSD)
gives a total  compliance cost of $60. Using this as the 2012MY cost  and applying time based
learning (as described in Chapter 3 of the Draft Joint TSD) results in a 2016MY cost of $53.

       In the 2008 Advance Notice of Proposed Rule, EPA presented a quick analysis of the
potential fuel savings associated with the control of indirect emissions via new AC
technology.  There EPA assumes  a reference 2010 fuel economy of 30 mpg for cars and 24
for trucks. With a 20% real-world shortfall, this becomes 24 and 19 mpg respectively.  As
described in appendix A of the GHG advanced notice (and above), A/C impacts overall fuel
consumption by 2.6-to-4.1%, and that an ultimate efficiency improvement of 40% is
achievable. EPA used the AEO 2008 fuel price, discount values, vehicle scrappage and VMT
figures employed  elsewhere in the advanced proposal to calculate a $96 cost savings for cars
and $130 for trucks forthe life of the vehicle. Assuming the same 0.23 factor to account for
rebound and emissions, these savings increase to $118 for cars and $159 for trucks. This was
noted in the GHG advance notice as being a potentially significant cost savings for the vehicle
owner compared to the cost of the efficiency improvements. EPA has not updated this
analysis for this rule. For the analysis in support of this rule, as presented in Chapter 6 of this
DRIA, the indirect AC fuel savings has been included in the total fuel savings resulting from
the proposal.

       Table 2-13 presents the compliance costs associated with new AC technology with
estimates for how those costs might change as vehicles with the technology are introduced
into the fleet. Costs shown are averages per vehicle since not all vehicles would include the
new technology but would, instead, include the technology according to the penetration
estimates shown in the table.

        Table 2-13: Estimated Costs in each Model Year for New AC Technology, 2007 dollars

Penetration
AC Leakage (Direct)
AC Indirect
Total
2012
25%
$4
$13
$18
2013
40%
$7
$21
$28
2014
55%
$9
$29
$39
2015
75%
$13
$40
$53
2016
85%
$14
$45
$60
J The 0.87 Euro-US dollar conversion is dated today but was valid in 2005. 2005 Euros are converted to 2005
US dollars then 2005 US dollars are converted to 2007 US dollars.
                                        2-38

-------
                                                                    Air Conditioning
2.5 Air Conditioning Credit Summary

       A summary table is shown with the estimated usage of the A/C credits.  EPA projected
the penetration rates as a reasonable ramp to the 85% penetration cap in 2016. The 85%
penetration cap was set to maintain consistency with the technology penetration caps used in
OMEGA.  The car and truck sales fractions were drawn from an adjusted version of AEO
2009, as documented in DRIA Chapter 5. As documented above, no use of alternative
refrigerant is projected in this in this analysis, although this assumption may be revisited in
the final rule (Table 2-14).

                 Table 2-14 : Credit Summary with Estimated Penetration Rates


Estimated Penetration
Car Sales Fraction
Truck Sales Fraction

Car Direct Credit
Car Indirect Credit
Total Car Credit

Truck Direct Credit
Truck Indirect Credit
Total Truck credit

Fleet average credits
Model Year
2012
25%
63%
37%

2.0
1.4
3.0

1.6
1.4
3.4

3.1
2013
40%
64%
36%

3.1
2.3
4.8

2.5
2.3
5.4

5.0
2014
60%
64%
36%

4.7
3.4
7.2

3.8
3.4
8.1

7.5
2015
80%
66%
34%

6.2
4.6
9.6

5.0
4.6
10.8

10. 0
2016
85%
66%
34%

6.6
4.8
10.2

5.4
4.8
11.5

10.6
                                        2-39

-------
Draft Regulatory Impact Analysis

        References

        All references can be found in the EPA DOCKET:  EPA-HQ-OAR-2009-0472.
1IPCC.  Chapter 2.  Changes in Atmospheric Constituents and in Radiative Forcing. September 2007.

2 Schwarz, W., Harnisch, J. 2003. "Establishing Leakage Rates of Mobile Air Conditioners." Prepared for the
European Commission (DG Environment), Doc B4-3040/2002/337136/MAR/C1.

3 Vincent, R., Cleary, K., Ayala, A., Corey, R. 2004. "Emissions of HFC-134a from Light-Duty Vehicles in
California." SAE 2004-01-2256.

4 EPA, 2009, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2007.
http://www.epa.gov/climatechange/emissions/usinventoryreport.html

5 State of California, Manufacturers Advisory Correspondence MAC #2009-01, "Implementation of the New
Environmental Performance Label," http://www.arb.ca.gov/msprog/macs/mac0901/mac0901.pdf.

6 State of Minnesota, "Reporting Leakage Rates of HFC-134a from Mobile Air Conditioners,"
http://www.pca.state.mn.us/climatechange/mac-letter-082908.pdf.

 Society of Automotive Engineers, "IMAC Team 4 — Reducing Refrigerant Emissions at Service and Vehicle
End of Life," June, 2007.

8 Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy. Transportation Energy Data
Book: Edition 27. 2008.

9 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2005.
http://epa.gov/climatechange/emissions/downloads06/07CR.pdf

10 EPA, "Fuel Economy Labeling of Motor Vehicle Revisions to Improve Calculation of Fuel Economy
Estimates Final Technical Support Document (179 pp, 2.1MB, EPA420-R-06-017,  December 2006

11 Ikegama, T., Kikuchi, K.. 2006. "Field Test Results and Correlation with SAEJ2727." Proceedings of the SAE
7  Alternative Refrigerant Systems Symposium.

12 Atkinson, W., Baker, J, Ikegami, T., Nickels, P. 2006. "Revised SAEJ2727: SAE Interior Climate Control
Standards Committee Presentation to the European Commission."

13 EPA.  2009 U.S. Greenhouse Gas Inventory Report. INVENTORY OF U.S. GREENHOUSE GAS
EMISSIONS AND SINKS: 1990-2007  (April 2009)

14 Society of Automotive Engineers, "IMAC Team 1 — Refrigerant Leakage Reduction, Final Report to
Sponsors," 2006.

15 Society of Automotive Engineers Surface Vehicle Standard J2727, issued August 2008, http://www.sae.org

16 Minnesota Pollution Control Agency, "Model Year 2009 Leakage Rate List,"
http://www.pca.state.mn.us/climatechange/mobileair.html.
                                              2-40

-------
                                                                               Air Conditioning
17 Schwarz, W., "Emission of Refrigerant R-134a from Mobile Air-Conditioning Systems," Study conducted for
the German Federal Environmental Office, September, 2001.

18 "A/C Triage — Ensuring Replacement Compressor Survival, " AC Delco Tech Connect, Volume 12, Number
5, January/February 2005,
http://www.airsept.com/Articles/CompressorGuard/ACDelcoTechConnectJanFeb05.pdf.

19 Johnson, Valerie H., "Fuel Used for Vehicle Air Conditioning: A State-by-State Thermal Comfort-Based
Approach," SAE 2002-01-1957, 2002.

20 Rugh, John P, Valerie Hovland, and Stephen O. Andersen, "Significant Fuel Savings and Emission Reductions
by Improving Vehicle Air Conditioning," Mobile Air Conditioning Summit, Washington DC., April 14-15,
2004.

21 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," 2004.

22 "Reducing Greenhouse Gas Emissions from Light-Duty Motor Vehicles," Northeast States Center for a Clean
Air Future, September 2004.

23 Fuel Economy Labeling of Motor Vehicles: Revision to Improve Calculation of Fuel  Economy Estimates;
Final Rule, 71 FR 77872, December 27, 2006.

24 Koupal,  J., "Air Conditioning Activity Effects in MOBILE6" EPA Report Number M6.ACE.001, 1998.

  EPA Technical Support Document for Fuel Economy Label Regulations, November, 2006.

26 Nam, E.K., "Understanding and Modeling NOx Emissions from Automobiles During  Hot Operation," PhD
Thesis, University of Michigan, 1999.

27 Nam, E. K., "Understanding and Modeling NOx Emissions from Air Conditioned Automobiles," SAE
Technical Paper Series 2000-01-0858, 2000.

28 "Reducing Greenhouse Gas Emissions from Light-Duty Motor Vehicles," Northeast States Center for a Clean
Air Future, September 2004.

29 Society of Automotive Engineers, "IMAC Team 2 - Improved Efficiency, Final Report," April 2006.

30 Memo to docket, "Meeting with Delphi and Presentation to EPA," March 2009.

31 Society of Automotive Engineers Surface Vehicle Standard J2765, issued October 2008, http://www.sae.org.

32 Barbat, Tiberiu, et al, "CFD Study of Phase Separators in A/C Automotive Systems,"  SAE 2003-01-0736,
2003.

33 EPA, "AC Idle Test Results_Summary for docket," August 2009.

34 "2009 Ward's Automotive Yearbook," Ward's Automotive Group, 2009. ISBN Number 978-0-9105-89-26-0.

35 California Air Resources Board, "Staff Analysis of Proposed Early Action for Climate Change Mitigation in
California," 2007, http://www.arb.ca.gov/cc/hfc-mac/documents/hfcdiy.pdf
                                               2-41

-------

-------
                                                     Technical Basis of the Standards
CHAPTER 3:  Technical Basis of the Standards

3.1   Technical Basis of the Standards

3.1.1  Summary

       As explained in section III.D of the preamble to the proposed rule, in developing the
proposed standard, EPA built on the technical work performed by the State of California
during its development of its statewide GHG program. This led EPA to evaluate a Clean Air
Act national standard which would require the same degree of technology penetration that
would be required for California vehicles under the California program. In essence, EPA
evaluated the stringency of the California Pavley 1 program but for a national standard.
However, as further explained in the preamble, before being able to do so, technical analysis
was necessary in order to be able to assess what would be an equivalent national new vehicle
fleet-wide CCh performance standards for model year 2016 which would result in the new
vehicle fleet in the State of California having CCh performance equal to the performance from
the California Pavley 1 standards.  This technical analysis is documented in this sub-chapter
oftheDRIA.

       Table 3-1 presents the calculated emission levels at which the national GHG standard
would ensure that vehicle sales in California of federally compliant vehicles would have fleet
average GHG emissions that are equal to the fleet average that would be achieved under the
California program described in Sections 1900, 1960 and 1961.1 of Title 13, California Code
of Regulations  ("Pavley I") by model year 2016:

         Table 3-1: Fleet Average National CO2 Emission Levels for Model Years 2012-2016

Fleet Average Tailpipe Emission Level
(CCh gram / mile)
MODEL YEAR
2012
288
2013
281
2014
275
2015
263
2016
250
       Manufacturer's use of credits and other program flexibilities may alter the program
stringency beyond that which is shown here.
                                        3-1

-------
Draft Regulatory Impact Analysis
3.1.2  Overview of Equivalency Calculation.

       The calculation of the fleet-wide national MY 2015 and MY 2016 CCh emission levels
which would be equivalent to California's Pavley I program is briefly outlined here.

    1.  Based on the California new vehicle fleet mix (predicted sales) and the CA program
       provisions, EPA calculated the fleetwide average CCh emissions achieved in CA from
       the 2015 and 2016 model year fleets.

    2.  The estimate of fleetwide average CCh emissions was disaggregated into achieved car
       and truck CCh emission levels at the national level using the new car and truck
       definitions proposed for this rule.

    3.  Based on the anticipated national fleet mix, the achieved car and truck levels were
       weighted together to determine the national targets which would achieve reductions
       equivalent to Pavley I in California.

    This calculation accounts for the compositional difference between the CA vehicle fleet
and the National fleet (i.e., CA has a higher proportion of cars than the average state), and for
various parameters in the CA program.

3.1.2.1   Calculating COi Equivalent Emissions under the California Program

       To calculate the CCh equivalent emissions  in California under Pavley I, the California
Passenger Car and Light Truck standards were combined with the California fleet mix in
order to calculate the anticipated emissions under the California standards from the California
fleet.

       The Passenger Car and Light Truck Standards were drawn from Sections 1900,  1960
and 1961.1 of Title 13, California Code of Regulations.  Intermediate and small volume
manufacturer standards were calculated based on guidance within the regulation, as well as
EPA analysis of current manufacturer product mix. These standards, less 2 grams per mile of
CO2 equivalent emissions due to methane (CH4) and nitrous oxide (N2O), are shown in Table
3-2. CH4 and N2O were excluded because the proposed EPA program separately addresses
these  emissions (Preamble  section III).
                                         3-2

-------
                                                      Technical Basis of the Standards
  Table 3-2: California Regulatory Standards excluding CH4 and N2O (grams CO2 equivalent per mile)

California Car (PC/LDT1) Standard
Intermediate/Small Volume
Manufacturer California Car Standard
CA LDT2/MDPV Standard
Intermediate/Small Volume
Manufacturer LDT2/MDPV Standard
MY 2015
Standard
211
314
339
360
MY 2016
Standard
203
229
330
357
       The projected fleet mix, as defined under Pavley I, was then determined in California.
Significantly, the California program deviates from historic definitions of "classic" cars and
trucks. In brief, Pavley I defines "PC/LDT1" as passenger cars and light duty trucks below
3,750 pounds, while "LDT2" include all trucks intended to convey passengers that weigh less
than 10,000 pounds.  The details of this classification scheme are found in the California
regulations.

       In order to estimate the emission contribution of PC/LDT1 and LDT2 in California,
EPA estimated the respective fleet fractions. EPA estimated the national sales mix in 2015
and 2016 at 60% passenger cars and 40% light duty trucks. This estimate is supported by the
Energy Information Administrations' Annual Energy Outlook 2009, which estimated
passenger cars at 59.4% of 2016 new vehicle sales in its published reference case.1 Due to the
American Recovery and Reinvestment Act of 2009, the Annual Energy Outlook reference
case has since been updated to project 2016 sales at 57.1% passenger cars.

       The projected 60% passenger cars, 40% light duty trucks sales fraction was then
applied to the California vehicle fleet mix. In such a scenario, the California Air Resource
Board (ARE) estimated that PC/LDTls comprise approximately 66% of the new light duty
vehicle fleet in California and that LDT2s comprise the remainder (34%).

       Once the PC/LDT1 and LDT2 fractions of California new vehicle sales were
determined, EPA estimated the fraction of vehicle sales in the intermediate and small volume
manufacturer categories. These manufacturers, which sell less than 60,000 vehicles per year
in California, are subject to less stringent emission standards under Pavley I.  While estimates
of future sales by manufacturer fluctuate, manufacturers such as Subaru, Porsche, Hyundai
and Volkswagen were considered beneath this threshold for the purpose of this analysis.
Based on EPA market analysis, small/intermediate volume manufacturers were estimated at
9% of total California PC/LDT1 sales and 5% of total California LDT2 Sales. The final
product mix assumed in California in 2015 and 2016 under a 60/40 national sales scenario is
shown in Table 3-3.

-------
Draft Regulatory Impact Analysis
  Table 3-3: California Sales Mix under a 60% Classic Car 40% Classic Truck National Sales Scenario

PC/LDT1 Sales
Intermediate Volume PC/LDT1
sales
California LDT2 Sales
Intermediate Volume LT2 sales
Sales %
60%
6%
32%
2%
       The product mix was multiplied by the relevant standard and summed in order to
calculate the achieved average CCh emissions for the new California fleet.  As an example in
2016:
Achieved Fleetwide CO2 Equivalent Emissions     =

(PC/LDT1 standard x PC/LDT1 Percentage) + (LT2 standard x LT2 Percentage) + (Intermediate Volume
PC/LDT1 standard x Intermediate Volume PC/LDT1 Percentage) + Intermediate Volume LT2 standard x
Intermediate Volume LT2 Percentage)
(0.6 x 203) + (0.06 x 229) + (0.32 x 330) + (0.02 x 357) = 248 grams.
                                                                                  (eq.l)
Based on the projected 60% passenger car, 40% light duty truck national sales mix (Table
3-3), the achieved fleetwide €62 equivalent tailpipe emission level expected in California in
2016 is 248 grams / mile.
       This analysis was repeated for model year 2015. In order to achieve equivalency, the
national program must produce a fleetwide average emission level in California that is no
higher than 261 grams CO2/ mile in 2015 and 248 grams CO2 / mile in 2016.
3.1.2.2        Translating the CA fleetwide average emissions into Cars (Passenger
        Automobiles) and Trucks (Non-Passenger Automobiles)

       In order to describe the national fleet, the California fleet-wide average CCh emission
level was translated into car and truck achieved emissions levels. However, the regulatory
definitions in EPA's Title II programs differ. Passenger Automobiles (PA) are defined as
two wheel drive SUVs below 6,000 Ibs. gross vehicle weight as well as classic cars.  The
remaining light duty fleet is defined as Non-Passenger Automobiles (NPA) (Table 3-4).
                                          3-4

-------
                                                    Technical Basis of the Standards
                      Table 3-4: Summary of Fleet Description Methods
REGULATOR
National Highway Transit
Safety Association (CAFE
Through MY 20 10)
California ARE
EPA
CAR DEFINITION
Car — Passenger Car

Car-PC + LDTl
Passenger Automobile — PC
+ 2 wheel drive SUVs below
6,000 GVW
TRUCK DEFINITION
Track - LOT 1 -4 and MDPV

Light Truck - LDT2-4 and MDPV
Non-Passenger Automobile — Remaining
light duty fleet
       To disaggregate the combined California fleet emission level into PA and NPA
vehicles, the 2015 and 2016 California achieved levels were multiplied by ratios derived from
National Highway Transit Association (NHTSA) analysis of the emissions from PA and NPA
vehicles.2  Based on the NHTSA analysis, EPA estimates that PAs have an emission
contribution equivalent to 91% of the California MY 2016 fleet average, while NPA have an
emission contribution equivalent to 119% of the California achieved CCh fleet average
emissions. These ratios, and the PA/NPA achieved emission levels, are shown in Table 3-5.

                    Table 3-5: PA and NPA Emission Levels under Pavley I
Regulatory Class



PA
NPA
Ratio



0.91
1.19
MY 2015
Achieved
Emission
Level
238
312
MY 2016
Achieved
Emission
Level
227
297
3.1.2.3  Calculating the 2015 and 2016 Fleetwide CO2 emission Targets under the EPA
        Proposal

       To determine the MY 2015 and MY 2016 fleetwide targets under the EPA proposal,
the achieved emission levels from PA and NPA (Table 3-5) were reweighted into a national
fleet-wide average based upon the anticipated national fleet of 60% passenger car,  40% light
duty truck. Based on NHTSA analysis presented in the MY 2011 CAFE final rule, this fleet
is expected to be comprised of approximately 66.4% PA and 33.6% NPA.3 The PA and NPA
achieved emission levels were weighted into a national fleetwide average based upon these
percentages. The resulting 2015 fleetwide target is 263 grams CCh / mile, while the 2016
target is 250 grams CO2 /mile.
                                        3-5

-------
Draft Regulatory Impact Analysis
3.1.2.4   Calculation of 2012-2014 "California Equivalent" Targets

       The methodology used to calculate the 2015 and 2016 California Equivalent levels
was repeated forthe 2012-2014 model years. The most significant departure from the
previously described methodology is that sales projections differ in MY 2012-2014 as
compared to MY 2015-2016.

       EPA assessment of projected vehicle sales during MY 2012-2014 supported a lower
proportion of car sales than the 60% fraction projected during MY 2015-2016.  March 2009
AEO vehicle sales estimates were therefore substituted in these earlier years. Using the
methodology described in section Error! Reference source not found., the AEO estimates
were used to project PC/LDT1 fractions in CA, and PA and NPA sales fractions nationally
(Table 3-6).
Comment [TShl]: Where is this
meant to point?
      Table 3-6: National PA and NPA Sales Fractions estimated in March 2009 AEO Projections
Regulatory Class
AEO Car fraction
AEO Truck fraction
PC/LDTlinCA
LT2 in CA
PA fraction Nationally
NPA fraction Nationally
MY 2012
55.0%
45.0%
61.0%
39.0%
62.1%
37.9%
MY 2013
56.1%
43.9%
62.1%
37.9%
63.0%
37.0%
MY 2014
57.4%
42.6%
63.4%
36.6%
64.1%
35.9%
       Per the previously described methodology, the calculated CA sales fractions were then
multiplied by the Pavley I standards for MY 2012 - MY 2014 (Table 3-7). Consistent with
the 2015/16 analysis, small manufacturers were assumed to remain a constant 9% of
California PC/LDT1 sales and 5% of California LDT2 Sales.

                        Table 3-7: 2012-2014 California Regulatory Standards
                       excluding CH4 and N2O (grams CO2 equivalent per mile)

California Car (PC/LDT1) Standard
Intermediate/Small Volume
Manufacturer California Car Standard
CA LDT2/MDPV Standard
Intermediate/Small Volume
Manufacturer LDT2/MDPV Standard
MY 2012
231
314
359
360
MY 2013
225
314
353
360
MY 2014
220
314
348
360
                                         3-6

-------
                                                     Technical Basis of the Standards
      The resulting achieved emission levels in California are 286 grams CO2 / mile in MY
2012, 279 grams CO2 / mile in MY 2013 and 273 grams CO2 / mile in MY 2014. In order to
derive PA and NPA achieved emission levels, these achieved emission levels were multiplied
by MY-specific ratios derived from National Highway Transit Association (NHTSA)
analysis.4

      The projected PA and NPA emission levels were then recombined into a national fleet
achieved emission level based on the national PA and NPA sales fractions shown in Table 3-6
(Table 3-8).
                    Table 3-8: PA and NPA Emission Levels under Pavley I
Regulatory Class
PA
NPA
Fleet Average
MY 2012
Achieved
Emission
Level
260
334
288
MY 2013
Achieved
Emission
Level
253
328
281
MY 2014
Achieved
Emission
Level
248
323
275
3.2  Analysis of Footprint Approach for Establishing Individual Company Standards

       One of the fundamental issues associated with the vehicle fleet average CO2emission
standard is the structure of the standard; i.e., the basis for the determination of the standard for
each vehicle manufacturer.

       Vehicle CO2 emissions are closely related to fuel economy.  Over 99 percent of the
carbon atoms in motor fuel are typically converted to tailpipe CO2, and therefore, for any
given fuel with a fixed hydrogen-to-carbon ratio, the amount of CO2 emitted (grams) is
directly correlated to the volume of fuel that is consumed (gallons), and therefore CO2 g/mile
is essentially inversely proportional to vehicle fuel economy, expressed as miles per gallon.
As part of the CAFE program, EPA measures vehicle CO2 emissions and converts them to
mpg and generates and maintains the federal fuel economy database. Additionally,  EPA
calculates the individual manufacturers' CAFE values each year, and submits  these  values to
NHTSA.

       EPA is proposing footprint-based CO2 standards for cars and light trucks. EPA
believes that this program design has the potential to promote CO2 reductions  across a broad
range of vehicle manufacturers, while simultaneously accounting for other important societal
objectives cognizable under section 202 (a) such as consumer choice and vehicle safety. EPA
believes a footprint-based system will also provide a more  level playing field among
manufacturers, as all models with similar size will have the same  CO2 emission targets, across
all manufacturers.

       In 2007, EPA evaluated several vehicle  attributes on which to base proposed CO2
standards for both cars and light trucks: footprint, curb weight, engine  displacement, interior
                                         3-7

-------
Draft Regulatory Impact Analysis
volume, and passenger carrying capacity. All of these attributes have varied advantages and
disadvantages. EPA's evaluation centered on three primary criteria (all of which reflect
factors relevant under section 202 (a)). 1) Correlation with tailpipe CCh emissions. Since
emissions of CO2 are controlled, there must be a reasonable degree of correlation from a
technical perspective between a proposed attribute and vehicle CCh emissions performance.
2) The relationship between the attribute and potential CChreducing technologies.  In order to
promote emissions reductions, choice in technology for the manufacturers, and cost-effective
solutions, it is important that an attribute not discourage the use of important CCh control
strategies.  3) How much the attribute would encourage compliance strategies that tend to
circumvent the goal of CCh reduction. EPA believes that it is important to choose an attribute
that minimizes the risk that manufacturers would change the magnitude of the attribute as a
method of compliance. 4) The  consistency of the attribute with existing or proposed
regulations. EPA does not want to create a program that competes with others that
accomplish similar goals. The 2007 analysis examines potential attributes against these
criteria and is outlined below.
3.2.1  "Footprint" as a vehicle attribute

       EPA is proposing to base the individual manufacturers fleetwide CCh standards on the
vehicle footprint attribute.  Footprint is defined as a vehicle's wheelbase multiplied by
average track width. In other words, footprint is the area enclosed by the points at which the
wheels meet the ground.

       In 2006, NHTSA adopted footprint as the basis for fuel economy standards in its
Reformed CAFE program for light trucks, and in 2008, the agency extended this program
structure to regulate passenger cars for MY 2011 and beyond. NHTSA used projected sales,
footprint, and mpg data from automakers' product plans, along with information on the cost
and effectiveness of fuel economy technologies, to create a footprint versus fuel economy
curve shown below in Figure 3-1  for cars and Figure 3-2 for trucks that establishes fuel
economy targets for every model's footprint value. Chapter V of NHTSA's RIA for the MY
2011 CAFE program contains more detailed information how the  MY 2011 car and truck
curves were generated.

-------
                                       Technical Basis of the Standards
        NHTSA Final MY 2011 Standards for Cars and Trucks
                   50        55        60
                         Footprint (ftz|
Figure 3-1 NHTSA Reformed CAFE Curve for MY 2011 Cars
                          3-9

-------
Draft Regulatory Impact Analysis
                         NHTSA Final MY 2011 Standards for Cars and Trucks
                40        45        50        55        60        65
                                         Footprint (ft2)



                 Figure 3-2 NHTSA Reformed CAFE Curve for MY 2011 Trucks
       The overall fleet-wide fuel economy compliance value for an individual manufacturer
is then calculated at the end of the model year by a sales-weighted, harmonic average of the
fuel economy targets for all models sold by that manufacturer. In the rulemaking process,
NHTSA also considered weight, towing capacity, and four wheel drive capability as
alternative attributes, but rejected them in favor of footprint.5

       EPA evaluated footprint as the attribute for setting vehicle CCh standards based on the
four criteria outlined above.
3.2.1.1   Correlation to tailpipe COi emissions

       Figure 3-3 and Figure 3-4 describe the relationship of tailpipe CCh emissions and
vehicle footprint. These figures were generated using the manufacturer's 2007 confidential
product plans, the most current projections at the time of the analysis.  EPA has since received
new product plans and developed a new baseline dataset from publicly available information.
However, EPA has not redone the analysis below with this new data as the general trends are
not expected to have changed.
                                         3-10

-------
                                                      Technical Basis of the Standards
       The first plot describes the model year 2007 car fleet and the second plot describes the
model year 2007 truck fleet. The circles represent the sales volume of a particular model,
where a larger circle corresponds to higher sales projection and a smaller circle corresponds to
a lower sales projection. In order to determine how closely footprint and CO2 emissions were
correlated, a linear least-squares regression was performed for cars and trucks separately.  It
should be noted that NHTSA used non-sales-weighted minimum absolute difference (MAD)
regressions to develop the slopes of the proposed fuel economy and CCh emission standards.
The reader is referred  to the preamble to the proposed rule for a discussion of the reasons for
use of non-sales-weighted MAD regressions for this purpose.
                                    44      46      48

                                          Footprint (ft2)
   Figure 3-3 Model Year 2007 Cars; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
                                       Footprint
                                         3-11

-------
Draft Regulatory Impact Analysis
                                                                       y=4.172x + 170.17
                                                                          R2 = .331
    3 450
    0
                                      55      60      65

                                          Footprint (ft2)
                                                            70      75       80      85
  Figure 3-4 Model Year 2007 Trucks; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
                                        Footprint
       As illustrated in the above figures, the R values for model year 2007 cars and trucks
are 0.283 and 0.331 respectively (both statistically significant to a confidence level greater
than 99%), indicating that there is a non-random correlation to CCh emissions. As vehicle
size increases, its CCh emissions tend to increase.
3.2.1.2   Relationship with CCh-reducing strategies

       The footprint attribute would encourage all CCh control strategies with the exception
of vehicle downsizing. All other things being equal, vehicle downsizing tends to correspond
to lower vehicle weight, which results in lower CCh emissions.  However, smaller vehicles
would have smaller footprints and would be subject to lower, more stringent, CCh emissions
targets,  discouraging downsizing as a compliance strategy. Also, absent other design
changes, decreasing vehicle size could reduce vehicle safety for that vehicle's driver,
especially for those vehicles less than 4000 pounds.6 Thus, the fact that footprint discourages
vehicle  downsizing is viewed by many safety advocates as a positive aspect.  This continues
to be an important factor in NHTSA's adoption of footprint in its Reformed CAFE program.
                                          3-12

-------
                                                     Technical Basis of the Standards
       A footprint attribute also would not discourage the use of lightweight materials, as a
lighter vehicle with no change in footprint would more easily comply with its €62 target.
Therefore, in choosing the footprint attribute, the use of lightweight material would remain a
viable compliance option, an important factor as lightweight materials can simultaneously
reduce mobile CC>2 emissions and improve vehicle safety. NHTSA came to the same
conclusion in its Reformed CAFE rulemaking.7  Though there can be a trend between weight
and size, EPA is not equating the two. Moreover, EPA is assuming that manufacturers can
and will lightweight their vehicles at a given footprint level as a potential compliance strategy.
3.2.1.3   Sensitivity of COi control to compliance-related vehicle adjustments

       Depending on the attribute, manufacturers may find it more economically attractive to
comply in a way that tends to compromise the expected emission reduction benefits of the
program.  Specifically, a manufacturer would have the opportunity to increase its average
fleet footprint over time in order to comply with a less stringent standard, which would
circumvent the CCh reduction goals of the program. However, major changes in a vehicle's
footprint typically require a substantial redesign of the vehicle, which typically occurs every
5-7 years.  NHTSA made this same finding in the Reformed CAFE rulemaking.8 While
definitive historical footprint data is not available, EPA believes that footprint has grown
more modestly in the past than many  other attributes.
3.2.1.4   Consistency with other existing or proposed regulatory programs

       EPA and NHTSA have coordinated closely in developing parallel GHG and MPG
standards in order to avoid creating a "patchwork" of regulations. Since NHTSA has in
recent history used footprint as the basis for its CAFE program and is proposing to continue
using this metric, footprint remains the simplest, most natural option with respect to the goal
of avoiding excessive regulatory burden on the manufacturers.

       Under the Clean Air Act, the State of California may petition EPA for the authority to
create more stringent mobile source emissions regulations at the state level. EPA has granted
California this privilege and the California program outlined does not utilize the footprint (or
any) attribute; instead the regulatory structure is based on a universal (or unreformed)
standard. Despite differences in the structure of the standards, the EPA federal program is
expected to have an equivalent stringency when compared to the California program, thus
making it a 50-state program. In order to account for early AC credits offered by the
California program, EPA has also chosen to adopt a very similar credit system outlined in
Chapter 2 of the RIA, which offers an additional layer of consistency.

-------
Draft Regulatory Impact Analysis
3.2.2  Alternative Attributes

       Some manufacturers have suggested using a vehicle's curb weight for an attribute-
based standard.  Curb weight is defined in EPA regulations (CFR 86.1803-01) as the actual or
estimated weight of the vehicle with all standard equipment,  plus the fuel weight at nominal
tank capacity, plus the weight of optional equipment. Figure 3-5 and Figure 3-6 below show
plots of tailpipe CCh emissions versus curb weight for 2007 car and truck models respectively,
where circle size indicates the sales volume of each model.
        2200    2400   2600    2800    3000    3200    3400    3600    3800    4000   4200    4400
                                         Curb Weight (Ibs)
Figure 3-5 Model Year 2007 Cars; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and Curb
                                         Weight
                                          3-14

-------
                                                       Technical Basis of the Standards
      650	'	^	o- -
                                                                      y = 0.0593x+126.85
                                                                          R2 = .521
      350	
        3000     3500     4000
                                4500      5000      5500      6000
                                         Curb Weight (Ibs)
  Figure 3-6 Model Year 2007 Trucks; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
                                      Curb Weight
       For both cars and trucks, curb weight has a relatively high correlation with tailpipe
CO2 emissions.  A sales-weighted linear least squares regression determined R2 values of
0.582 for cars and 0.521 for trucks, indicating a substantial relationship of the current fleet's
curb weight and CCh emissions.

       Historically, some vehicle safety advocates have preferred weight for an attribute-
based standard since a standard with a steep relationship with weight discourages down-
weighting. However, with recent advances in strong, lightweight materials, occupant safety is
not necessarily compromised by a reduction in vehicle weight.9 In fact, these studies have
shown that a vehicle's size is a more important factor than weight in its effect on occupant
safety. In a weight-based attribute system, a lower weight would correspond to a more
stringent CCh standard.  While this would discourage downsizing as a compliance strategy,
it's important to recognize that weight as an attribute for determining tailpipe  CCh standards
would discourage the use of lightweight materials, even though advanced lightweight
materials could simultaneously reduce CCh emissions and improve vehicle safety.

       Furthermore, since a vehicle's weight is much easier to change than most other
attributes, it is more likely that manufacturers could add weight to their vehicles in order to be
subject to and comply with a  less stringent standard.  This potential is reinforced by the
relatively high rate of growth of vehicle weight; it has grown  1.0 - 1.5% per year since the
                                          3-15

-------
Draft Regulatory Impact Analysis
late 1980s.10  This development would have negative environmental consequences by
increasing overall €62 emissions, contrary to the chief goal of section 202 (a) of the Act.

       EPA also examined engine displacement as a potential attribute for determining
manufacturer CCh standards.  Engine displacement is defined as the volume swept as the
piston moves from top dead center to bottom dead center.  Figure 3-7 and Figure 3-8 below
contain sales-weighted linear regression plot of tailpipe CCh emissions and engine
displacement for 2007 cars and trucks, respectively.
   1
   S 300
   o
   o
     150	Q=
                             2345
                                      Engine Displacement (L)
  Figure 3-7 Model Year 2007 Cars; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
                                  Engine Displacement.
                                         3-16

-------
                                                       Technical Basis of the Standards
    3 450
    0
                                                                       y=45.094x +217.45
                                                                          R2 = .619
                                       Engine Displacement (L)


  Figure 3-8 Model Year 2007 Trucks; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
                                   Engine Displacement
       Engine displacement correlates well to tailpipe emissions, with R values of 0.667 for
cars and 0.619 for trucks.  This is because increasing engine displacement typically increases
the amount of fuel burned per cycle.

       EPA believes that a standard based on engine displacement does not guarantee any
environmental benefit because of the disincentive to add certain CCh-reducing technologies
and the potential for manufacturers to adjust the sales of higher-displacement models
regardless of whether or not it reflects market demand.   Hypothetically, a model could have
three trim lines with three different displacements: A 4-cyUnder 2.0L Turbo, a 4-cylinder
2.5L, and a 6-cylinder 3.0L. Since these models would have three standards ranging from
most to least stringent, correspondingly, this type of standard would be a disincentive to sell
models with smaller engines or turbochargers.  These strategies can dramatically reduce CCh
emissions (See RIA Section on Tech Feasibility) and are increasingly prevalent in the
European market. Thus EPA believes that the use of engine displacement for establishing
CO2 tailpipe standards will undermine readily achievable and feasible reductions of CCh
emissions.
                                          3-17

-------
Draft Regulatory Impact Analysis
       EPA also examined interior volume and occupant capacity as potential attributes
because they characterize vehicle utility well.  Increasing interior volume creates more space
for people and cargo, and increasing occupant capacity creates the potential to carry more
people, both important factors consumers consider when purchasing a new vehicle. Figure
3-9 below contains a plot of interior volume and tailpipe CCh for model year 2007 cars.
                       MY 2007 Cars: Tailpipe CO2 Emissions by Interior Volume
                                gffifffia.rr   ::  •
                                      110        130

                                       Inter!or Volume (H3|
  Figure 3-9 Model Year 2007 Cars; Linear Trend of CO2 Tailpipe Emissions and Engine Displacement
       EPA confirmed that interior volume is not at all correlated to vehicle CCh emissions
with a R2 value of 0.0036 for cars.  The correlation of interior volume and tailpipe CO2 is
worse for light trucks by definition, since cargo space for pickup trucks is a separate exterior
bed. Thus, it does not make sense to have a €62 standard for light trucks that is based on
interior volume, since pick-up trucks would be required to meet a stricter CO2 standard than
SUVs and minivans, which are typically regulated in the general "truck" category. For these
reasons, EPA is not proposing interior volume for the standard.

       Alternatively, occupant capacity does not share the same safety implications as
interior volume. Furthermore, since it is difficult to game and does not discourage the use of
any CCh -reducing technologies, there is significant potential for CCh improvement.  Figure
3-10 and Figure 3-11 below illustrate the breakdown of the model year 2007 fleet in terms of
occupant capacity.
                                        3-18

-------
                                                Technical Basis of the Standards
              Model Year 2007 Cars: Percentage of Sales by Occupant Capacity
                                   Number of Seats
Figure 3-10 Model Year 2007 Cars; Percentage of Sales by Occupant Capacity
                                  3-19

-------
Draft Regulatory Impact Analysis
                       MY 2007 Cars: Range of Tailpipe CO2 Emissions by Occupant Capacity
                                           Number of Occupants

      Figure 3-11 Model Year 2007 Cars; Range of Tailpipe CO2 Emissions by Occupant Capacity

       However, occupant capacity and CO2 emissions do not relate well.  Since 84% of the
2007 car fleet has 5 seats, an occupant-based standard would essentially result in a universal
standard for a majority of vehicles.  Since the car models falling into the 5-seat category have
a tailpipe €62 range of 133 to 472 g/mi, an occupancy-based standard would negate the
benefits from relative equity of the attribute-based system to full line manufacturers.

3.2.3  EPA Selection of the Footprint Attribute

       EPA has considered a range of potential vehicle attributes that could be used to set
CO2 standards. To summarize key results from the 2007 analysis, interior volume  and
passenger carrying capacity have extremely poor correlation with fuel economy, and EPA is
not proposing them for that reason.  The three remaining attribute options—footprint, curb
weight, and engine displacement—are all reasonable choices in terms of correlation with CCh
emissions levels, with weight having the best correlation to CCh emissions levels.  However,
it should be noted that correlation is not the primary deciding factor for the selection of an
attribute.  One could easily get an excellent correlation by choosing a function that combines
the effects of weight, displacement,  N/v ratio (engine speed  to vehicle speed ratio at top gear),
and frontal area (as a product with the aerodynamic coefficient). There are many other, but
these are the four variables that most define a vehicle's fuel  economy11'12  The choice of an
attribute  is not only an engineering decision, it also a policy decision. It is linked with the
outcomes that are desired in a future fleet.
                                          3-20

-------
                                                      Technical Basis of the Standards
       With respect to the remaining criteria, EPA believes footprint is clearly superior to
both weight and engine displacement. Footprint does not inherently discourage any key €62
control strategies (except for vehicle downsizing), while weight would discourage the use of
lightweight materials.  Engine displacement would discourage engine downsizing with
turbocharging, a strategy increasingly popular in the United States and Europe. Footprint is
somewhat less susceptible to modifications for compliance, since major changes would
generally require a significant platform redesign; in contrast, it is easier for manufacturers to
change weight and engine displacement.

       EPA notes that the footprint attribute also correlates well with the "utility" or
"usefulness" of the vehicle to the consumer. Larger footprints amount to more space inside
the vehicle to carry passengers or cargo, which are important considerations for consumers.
Thus, it is an additional benefit that the footprint-based approach would not discourage
changes to vehicle designs that can provide more utility to consumers. EPA also recognizes
that if footprint is used for the vehicle €62 standards then the form of the standards would be
compatible with NHTSA's use of footprint in their Reformed CAFE program.  EPA requests
comment on the proposed selection of the footprint attribute for establishing manufacturer-
specific CO2 standards.

       For these reasons, EPA therefore believes that the footprint attribute is the best choice
of the attributes discussed, from both an engineering as well as from a public policy
standpoint. EPA therefore proposes to use footprint in the CO2 standard-setting process for
this rule.

       EPA is proposing to implement the footprint attribute in the proposed CO2 control
program via a piecewise linear function.  As mentioned above, this is the equivalent to the
shape selected by NHTSA for its proposed CAFE standards for light trucks for model years
2012-2016.  The shape of this function with respect to CO2 is  reflected in Figures I.D.3-3 and
I.D.3-4 of the preamble. The difference is that it moves from low CO2 values on the left to
high CO2 values on the right (see Figure 3-12 and Figure 3-13 below for example) due to its
inverse relation to MPG.
                                         3-21

-------
Draft Regulatory Impact Analysis
     400-
     350 -
   -i
   'E
    |300-
   o
   o
     250 -
     200
        30
     400-
     350 -
   -i
   'E
   s
   O
     250-
     200
        30
                2012
               2016
                  35
                2012
                2016
                  35
                             40
                                           Footprint (sq feet)



                            Figure 3-12 CO2 (g/mi) Car standard curves
                             40
                                       45  _  .  .50.  ,  ..  55
                                           Footprint (sq feet)
                                                                      60
                           Figure 3-13 CO2 (g/mi) Truck standard curves
                                                                                65
                                                                                          70
                                                                                65
                                                                                          70
                                              3-22

-------
                                                     Technical Basis of the Standards
Implementing the proposed CO2 emission standards in this manner provides consistency with
NHTSA's proposed CAFE standards.

       Other forms of a footprint-based CO2 standard are possible. Examples are a logistic
curve, simple straight line, a line which levels out at a certain point to discourage vehicle
upsizing, etc. Section II of the preamble as well as Chapter 2 of the draft joint TSD contains
more information on how EPA defined the piecewise linear CO2 target function.
3.3  Supplemental Analysis of Relative Car and Truck Standards

       The methodology used to set the standards, and thus, the relative stringency of the car
and truck standards is described in Chapter 2 of the draft Joint TSD.  The car and truck
standards were set based on achieved levels of car and truck fuel economy from NHTSA's
Volpe Model under conditions where estimated net social benefits were maximized, while the
application of strong hybrid and diesel technology was excluded. These achieved levels were
then adjusted upwards arithmetically until the combined car-truck fleet met the CO2 levels
described in Section 3.1 above. EPA and NHTSA worked jointly in that analysis and EPA
believes it is a reasonable basis for determining the relative car-truck stringency for the
proposed GHG standards.

       In this section, a few alternative methods to setting this relative stringency are
investigated. These  methods use the OMEGA model, though it is expected that the results
would be similar if the Volpe Model was used.  EPA's OMEGA model and its use in support
of this proposed rule is described in Chapter 4 of this draft RIA.

       In performing these runs of the OMEGA model, the technology packages were
removed which added either hybrid or diesel technology.  This was done because the CO2
reduction required to meet the  overall level of CO2 emissions being proposed for 2016, 250
g/mi, is generally not requiring significant levels of either hybrid or diesel technology when
the potential for A/C credits is considered.  This is the same approach that was used in the
Volpe modeling used to develop the car and truck standards as described in Chapter 2 of the
draft Joint TSD.  The technology penetration of the other technologies was limited as
described in Chapter 1 for MY 2016; 100% for the technologies included in package 1 for
each vehicle type and 85% for the more significant technologies (any technology package
number 2 or higher). In order to be comparable to the Volpe modeling, A/C related
technologies were not included.  Thus, the CO2 emission target was 261 g/mi in 2016.  Fuel
prices and the value  of externalities were the same as those used to project the benefits from
the proposal.

       As described in greater detail in Chapter 4, the OMEGA  model applies technology
incrementally. It determines which vehicle receives the next step of control using the
manufacturer-based  net cost effectiveness metric. This metric includes the cost of the
technology, the value of the fuel savings which accrues from the CO2 technology over a
specified time period (here, 5 years) and the degree of CO2 emission reduction.  This ranking

-------
Draft Regulatory Impact Analysis
process includes the decision of whether to apply the next level of control to a car or a truck.
This allows the evaluation of car and truck CO2 control simultaneously.

       In the first set of OMEGA runs used to evaluate relative car-truck stringency, all
manufacturers' vehicles were combined into a single fleet. Determining manufacturer-
specific CO2 standards or levels of control requires some initial assumption of relative car
and truck stringency. Modeling the industry as one, single fleet avoids this. It also makes the
analysis more technical in nature and focuses on the technical capability of the industry to
reduce car and truck CO2 emissions efficiently, as manufacturer distinctions are eliminated.
The disadvantage is that it does not reflect the fact that manufacturers vehicles differ in their
fundamental level of CO2 emissions and therefore, manufacturers differ in the projected level
of technology needed to meet the same footprint-based standard.  With this simplification, a
single run of the OMEGA model can provide an indication of the relative potential to control
car and truck CO2 emissions.

       We ran the OMEGA model with a CO2 emission standard which was beyond the
capability of the technology available (i.e., maximum technology).  The combined level of car
plus truck CO2 emissions was determined, as well as for cars and trucks separately.  The
overall level of control was relaxed until net societal benefits were maximized. The benefits
quantified included reduced externalities related to crude oil imports, upstream emissions,
refueling time and CO2 emission reductions.  They also included costs and benefits associated
with increased VMT due to the fuel economy rebound effect (e.g., vehicle emissions, noise,
congestion, value of additional driving, etc.)  CO2 was valued at $20 per ton in 2007 and a
3% per annum discount rate was used. Finally, the overall CO2 emissions was increased in
several steps from the  maximum net benefit level to evaluate how relative car and truck
emissions changed with the overall level of stringency.  The results of this analysis are
summarized in Table 3-9.

 Table 3-9 Relative Car and Truck CO2 Emissions At Various Levels of CO2 Control - Single Industry-
                     wide 2016 Fleet (Excludes Impact of A/C Technology)

Maximum Application of
Technology
Maximum Net Societal Benefits
+5 g/mi CO2
+ 10g/miCO2
+ 15g/miCO2
+20 g/mi CO2
+25 g/mi C02
Car CO2
(g/mi)
209.4
210.1
214.8
218.6
222.0
222.2
226.0
Truck CO2
(g/mi)
286.6
286.6
291.7
298.3
305.5
318.5
325.2
Combined CO2
(g/mi)
235.3
235.8
240.6
245.3
250.0
254.5
259.2
Ratio of Truck/Car
CO2
1.37
1.36
1.36
1.36
1.38
1.43
1.44
       As can be seen, applying all the non-hybrid and non-diesel technology achieved an
overall level of CO2 control of 235 g/mi. This is equivalent to a fuel economy of 37.8 mpg,
since the use of diesel engines in only that in the reference fleet. (If A/C-related CO2
emission reduction of 11 g/mi were subtracted from this figure, it would decrease to 224 g/mi,
equivalent to 3 9.6 mpg.) The ratio of car to truck CO2 emissions at this overall level of CO2
emissions is 1.37. Relaxing this level until net societal benefits reach their maximum only
increased CO2 emissions by 0.5 g/mi (equivalent to a decrease in fuel economy of 0.1  mpg).
                                         3-24

-------
                                                      Technical Basis of the Standards
This indicates that net incremental cost of nearly all the non-hybrid, non-diesel technology
packages are less than the societal benefits provided.  The ratio of car to truck CO2 emissions
at this overall level of CO2 emissions decreases slightly to 1.36.  The ratio for the proposed
truck and car standards in 2016 is 1.35 (302/224), or nearly identical.

       As the overall level of CO2 control is relaxed, the ratio of truck to car emissions
increases.  This indicates that, on the increment, the least cost effective steps of control were
applied primarily to trucks and not cars. EPA's and NHTSA's methodology for setting the
relative stringency for the car and truck standards in 2012 through 2016 began with the car
and truck CO2 levels where estimated social benefits were maximized and increased the CO2
levels of both cars and trucks by the same increment until the combined CO2 level was 261
g/mi. Because the cars emit less than trucks, these shifts upward increase car emissions by a
greater percentage than truck emissions.  Increasing the overall level of CO2 emissions based
on the relative cost effectiveness of technology would increase truck emissions by a greater
percentage than car emissions.
                                         3-25

-------
Draft Regulatory Impact Analysis
        References

        All references can be found in the EPA DOCKET: EPA-HQ-OAR-2009-0472.
1 Energy Information Administration. Annual Energy Outlook 2009.
http://www.eia.doe.gov/oiaf/aeo/index.html

2NHTSAModel Year 2011 Rule. RIN 2127-AK29. Average Fuel Economy Standards.  Passenger Cars and
Light Trucks. Model Year 2011.

3 NHTSA Model Year 2011 Rule. RIN 2127-AK29. Average Fuel Economy Standards.  Passenger Cars and
Light Trucks. Model Year 2011.

4 NHTSA Model Year 2011 Rule. RIN 2127-AK29. Average Fuel Economy Standards.  Passenger Cars and
Light Trucks. Model Year 2011.

5 See generally 71 FR at 17595-96.

6 National Academy of Sciences, "Effectiveness and Impact of Corporate Average Fuel Economy (CAFE)
Standards," National Academy Press, Washington, DC, 2002. Available for online viewing or hard copy
purchase from the National Academy Press at http://books.nap.edu/openbook.php?isbn=0309076013.

7 71 FR at 17620-21; see also 2002 NAS Report at 24.

8 71 FR at 17613.

9 71 FR at 17596; 2002 NAS Report at 24.

10 Light-Duty Automotive Technology and Fuel Economy Trends:  1975 through 2007," U.S. Environmental
Protection Agency, EPA420-S-07-001, September 2007, "http://www.epa.gov/otaq/fetrends.htm

11 Nam, E.K., Giannelli, R, Fuel Consumption Modeling of Conventional and Advanced Technology Vehicles
in the Physical Emission Rate Estimator (PERE), EPA document number EPA420-P-05-001, 2004

12Guidelines for Analytically Derived Fuel Economy. March 11, 2004
http://www.epa.gov/otaq/cert/dearmfr/ccd0406.pdf
                                             3-26

-------
                                    Results of Proposed and Alternative Standards

CHAPTER 4:  Results of Proposed and Alternative Standards

4.1 Introduction

       There are many ways for a manufacturer to reduce CC>2 emissions from any given
vehicle during a redesign. 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 CC>2 emissions control, there are an almost
infinite number of technology combinations which produce the desired CC>2 reduction.
EPA has created a new vehicle model, the Optimization Model for Emissions of
Greenhouse gases from Automobiles (OMEGA) 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.

4.2  Model Inputs

       OMEGA 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.  Chapter 1 of the TSD contains a description
of how the vehicle reference fleets were created for modeling purposes, and includes a
discussion on how EPA defined the  19 vehicle types. In addition, the degree to which
each vehicle already reflects the effectiveness and cost of each available technology in
the 2008 baseline fleet must also be input.  This prevents the model from adding
technologies to vehicles already having these technologies in the baseline. It also avoids
the situation, for example, where the model might try to add a basic engine improvement
to a current hybrid vehicle. Section 4.2.1 of this Draft Regulatory Impact Analysis
(DRIA) contains a detailed discussion of how EPA accounts for technology present in the
baseline fleet in OMEGA.

       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 which determine the individual technology cost and effectiveness
values are not explicitly used by the model; instead, the costs and effectiveness used by
the model are associated with each vehicle package, and are based on their associated
vehicle types (of 19). This information was described in Chapter 1 of this DRIA and
Chapter 3 of the Draft Joint TSD. In all cases, the order of the technologies or
technology packages for a particular vehicle type is designated by the model user in the
input files prior to running the model.  Several criteria can be used to develop a
reasonable ordering of technologies or packages.  These are described in Chapter 1 of the
Draft RIA.

       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  chapter 4 of the Draft Joint TSD.
                                       4-1

-------
Draft Regulatory Impact Analysis

       The fourth type of data describes the CC>2 emission standards being modeled.
These include the CC>2 emission equivalents of the 2011 MY CAFE standards and the
proposed CO2 standards for 2016. As described in more detail in Chapter 2 of this DRIA
and briefly in section 4.2.1 below, the application of A/C technology is evaluated in a
separate analysis from those technologies which impact CC>2 emissions over the 2-cycle
test procedure. For modeling purposes, EPA applies this AC credit by adjusting
manufacturers' car and truck CO2 targets by an amount associated with EPA's projected
use of improved A/C systems, as discuss in Section 4.2.1, below.

4.2.1 Representation of the CO2 Control Technology Already Applied to
      2008 MY Vehicles

       The market data input file utilized by OMEGA, which characterizes the vehicle
fleet, is designed to account for the fact that vehicles may be equipped with one or more
of the technologies available in general to reduce CO2 emissions. As described in
Chapter 1 of this RIA, EPA decided to apply technologies in packages, as opposed to one
at a time.  However, 2008 vehicles were equipped with a wide range of technology
combinations, many of which cut across the packages.  Thus, EPA developed a method to
account for the presence of the combinations of applied technologies in terms of their
proportion of the EPA packages described in Chapter 1. This analysis can be broken
down into four steps.

       The first step is to develop a list of individual technologies which are either
contained in each technology package, or would supplant the relevant portion of each
technology package  (e.g., the engine, the transmission,  etc.). For example, variable
intake valve timing would be associated with a downsized, turbocharged, direct injection
engine. Thus, the effectiveness and cost of variable intake valve timing would be
considered to be already present for any technology package which included either
variable intake valve timing or included an engine technology which provided greater
effectiveness. The reverse case would be an example of a technology which would
supplant another technology. If a vehicle already had a downsized, turbocharged,  direct
injection  engine, the effectiveness and cost of this technology would be considered to be
already fully present when evaluating the application of a technology package which
included variable intake valve timing or any other engine technology up to and including
a downsized, turbocharged, direct injection engine.  In either case, the effectiveness and
cost present on the 2008 MY vehicle would be limited to the effectiveness and cost of the
engine technology This would allow the application of non-engine related technologies
also included in that package (e.g., an improved transmission) to still be applied.

       A specific example would be a 2008 MY vehicle falling into EPA vehicle type  1
equipped with a dry  DCT. A dry DCT is added as part of EPA's technology package
number 3 for vehicle type 1. Thus, the effectiveness and cost of a dry DCT would be
considered to be already present when applying the effectiveness and cost of package 3 to
this vehicle. If the dry DCT contributed to 50% of the total effectiveness and 40% of the
cost of package 3 over package 2 and 40% of the cost of package 3 over package 2 for
this vehicle, then these percentages would be included in the market data file for this
vehicle. If the level  of CO2 control led to the application of technology package 3 to this

                                       4-2

-------
                                    Results of Proposed and Alternative Standards

vehicle in a run of the OMEGA model, the model would only apply 50% of the
effectiveness of package 3 to this vehicle and 60% of the cost of this package.

       If two consecutive technology packages both contain a dry DCT, packages 3 and
4, for example, the model user does not need to declare it as technology already present
in the baseline. Instead, the benefit of package 4 is due to engine-related technology is
considered incremental to package 3 and is contained in the technology input file.  When
applying package 4 to this vehicle, the OMEGA model will apply its full incremental
effectiveness and cost.  The same would be true for packages which replace a  dry DCT
with a hybrid technology 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.

       Since vehicle  models are grouped into vehicle types for technology addition, and
there are relatively few vehicle types compared to the number of vehicle models
analyzed, there may be special cases when the applicable technology packages are less
effective than baseline technologies; EPA therefore limits the applicability of the relevant
technology packages  to the affected vehicle models by adjusting the cost and
effectiveness reflected in the baseline by the same method described above. An example
of a single 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  small improvements to
a gasoline engine, since the resultant gasoline engine would otherwise accrue technology
packages which are not as effective at reducing CO2 than the baseline 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.

       Six technologies were considered to be  contained in all technology packages
available for all 19 vehicle types: low viscosity lubrication, improved engine friction, low
rolling resistance tires, high efficiency accessories, electric power steering and improved
automatic transmission controls and lock-up. The total effectiveness and cost estimates
for every technology  package described in Chapter 1 includes the  effectiveness and cost
of these five technologies.  Thus, their presence on a 2008 MY vehicle must be deducted
from the effectiveness of any package added in the modeling.

       All the packages described in Chapter 1 which  are numbered 2 or higher (i.e., all
packages other than the initial package available to each vehicle type) contain at least
dual cam phasing and a 6-speed automatic transmission.  Thus, the followed technologies
were considered to be contained in all technology packages numbered 2 and above for all
19 vehicle types: intake valve timing, coupled cam phasing, dual cam phasing, 6-speed
(or higher) automatic or manual transmission, or a continuously variable transmission. In
addition, the following technologies supplant either dual cam phasing or a 6-speed

                                       4-3

-------
Draft Regulatory Impact Analysis
automatic transmission: variable valve lift and timing, cylinder deactivation, diesel
combustion, dual clutch transmission, and IMA, power-split and 2-mode hybridization.
Thus, the effectiveness and cost of these technologies were also included when
determining the percentages of the effectiveness and cost of each package number 2 or
higher which was already utilized on each 2008 MY vehicle.

       All the packages described in Chapter 1 which are numbered 2 or also include
conversion to direct injection gasoline combustion, except for package 2 for vehicle type
1. Thus, the effectiveness and cost of direct injection gasoline combustion was also
included when determining the percentages of the effectiveness and cost of each package
number 2 or higher which was already utilized on each 2008 MY vehicle, with the
exception of package number 2 for vehicle type 1.

       Table 4-1 depicts the packages which first turbocharge an engine.

            Table 4-1 Technology Packages Which Initially Include Turbocharging
Vehicle
Type
1
3
4
5
6
7
8
9
10
Technology
Package
4
4
2
4
4
5
5
4
5
Vehicle
Type
12
13
14
15
16
17
18
19

Technology
Package
5
5
5
4
3
5
5
5

       Thus, the effectiveness and cost of turbocharging was included when determining
the percentages of the effectiveness and cost of each package shown in Table 4-1 or
higher within each vehicle type which was already utilized on each 2008 MY vehicle.

       Table 4-2 depicts the packages which first add start-stop technology to a vehicle.
                                       4-4

-------
                                    Results of Proposed and Alternative Standards
               Table 4-2 Technology Packages Which Initially Add Start-Stop
Vehicle
Type
1
2
3
4
5
6
7
8
9
10
Technology
Package
3
3
3
3
3
3
4
4
4
4
Vehicle
Type
11
12
13
14
15
16
17
18
19

Technology
Package
4
4
4
4
3
3
4
4
4

       Thus, the effectiveness and cost of start-stop technology was included when
determining the percentages of the effectiveness and cost of each package shown in Table
4-2 or higher within each vehicle type which was already utilized on each 2008 MY
vehicle.

       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.

       The fourth step is to combine the fractions of the cost and effectiveness of each
technology package already present on the individual 2008 vehicles models for each
vehicle type. 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 vehicle type. 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 vehicle type. Once
again, this process prevents the model from adding technology which is already present
                                       4-5

-------
Draft Regulatory Impact Analysis

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 proposed CO2 standards.

       Table 4-3 and Table 4-4 show the degree to which the baseline fleet, adjusted for
sales in 2016, includes the effectiveness and cost of the various technology packages by
vehicle type.

  Table 4-3 Presence of Technology on 2008 MY Vehicles In Terms of CO2 Effectiveness (Weighted
                      Average Across Car and Truck Sales in 2016)

Vehicle
Type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Technology Package Number
1
6.50%
13.20%
11.80%
2.80%
11.40%
2.20%
8.30%
0.00%
0.50%
0.50%
1.40%
7.50%
22.90%
0.00%
0.00%
5.60%
4.00%
19.40%
17.80%
2
19.70%
28.10%
21.70%
35.30%
38.30%
47.70%
22.10%
0.00%
0.00%
4.70%
0.00%
13.00%
6.80%
33.70%
0.00%
31.30%
57.70%
15.80%
47.40%
3
2.30%
12.60%
0.50%
0.00%
3.70%
1.60%
19.40%
0.00%
0.00%
0.60%
0.00%
2.40%
57.40%
0.00%
0.00%
3.30%
9.40%
0.00%
0.00%
4
0.00%
N/A
24.30%
0.00%
1.00%
4.10%
0.10%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
5
N/A*
N/A
N/A
N/A
N/A
N/A
0.00%
0.00%
0.00%
0.00%
N/A
0.00%
0.00%
0.00%
0.00%
N/A
0.00%
0.00%
0.00%
6
N/A
N/A
N/A
N/A
N/A
N/A
0.00%
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
 : N/A: No such package for that vehicle type
                                       4-6

-------
                                    Results of Proposed and Alternative Standards
Table 4-4 Presence of Technology on 2008 MY Vehicles In Terms of Cost (Weighted Average Across
                             Car and Truck Sales in 2016)

Vehicle
Type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Technology Package Number
1
1.40%
7.60%
8.10%
2.00%
10.90%
0.70%
6.20%
0.00%
0.10%
0.10%
0.20%
1.30%
3.30%
0.00%
0.00%
1.80%
1.40%
12.60%
2.70%
2
16.50%
23.20%
15.40%
31.90%
22.40%
33.80%
19.90%
0.00%
0.00%
2.60%
0.00%
3.10%
2.80%
8.70%
0.00%
22.90%
46.90%
11.10%
26.60%
3
1.90%
16.40%
3.40%
0.00%
3.40%
2.40%
27.90%
0.00%
0.00%
0.60%
0.00%
2.40%
58.20%
0.00%
0.00%
6.90%
12.00%
0.00%
0.00%
4
0.00%
N/A
5.70%
0.00%
0.20%
3.00%
0.20%
0.00%
0.00%
0.00%
0.00%
0.00%
0.10%
0.00%
0.00%
0.00%
9.50%
0.00%
0.00%
5
N/A
N/A
N/A
N/A
N/A
N/A
0.20%
0.00%
0.00%
0.00%
N/A
0.00%
0.10%
0.00%
0.00%
N/A
0.00%
0.00%
0.00%
6
N/A
N/A
N/A
N/A
N/A
N/A
0.00%
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
       As mentioned above for the market data input file utilized by OMEGA, which
characterizes the vehicle fleet, the 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 the Draft TSD Chapter 3. Because EPA chose 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.
                                       4-7

-------
Draft Regulatory Impact Analysis

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.

4.2.2 Technology Package Approach

        Consistent with its streamlined redesign cycle approach, EPA designed OMEGA
to allow the user to add GHG-reducing technologies in packages that would reasonably
and likely be added by manufacturers within a redesign cycle. In addition, the user can
combine similar vehicle models into "vehicle type"  groups which are likely to receive the
same list of technology packages. For each vehicle  type, the user must rank the
technology packages in order of how OMEGA should add them to that specific vehicle
type.  This approach puts some onus on the user to develop a reasonable sequence of
technologies. However, the model also produces information which helps the user
determine when a particular technology or bundle of technologies might be "out of
order".  The approach also simplifies the model's calculations and enables synergistic
effects among technology packages to be included to the fullest degree possible.

             When technology is sufficiently new, or the lead time available prior to
the end of the redesign cycle is such that it is not reasonable to project that the technology
could be applied to all vehicle models that are of the same specific vehicle type, the user
can limit the technology application through the use of a market penetration cap ("market
cap") of less than 100%.  This cap can vary by redesign cycle.  When a technology
package is applied to fewer than 100% of the sales of a vehicle model due to the market
cap, the effectiveness of the technology group is simply reduced proportionately to reflect
the total net effectiveness of applying that technology package to that vehicle's sales.
Most of the technologies for the analysis conducted in this rule had a market cap of 85%,
though hybrids were restricted to 15% for reasons described in the preamble. A small
number of technologies had a 100% phase in cap. These include: low friction lubricants,
electric power steering, improved accessories, and low rolling resistance tires.  These
simple to apply technologies may be implemented outside of a vehicle's normal redesign
schedule.

             OMEGA does not create  a new vehicle with the technology package and
retain the previous vehicle which did not receive the technology package, splitting sales
between the old and new vehicles. If subsequent technology packages can be applied to
the vehicle, the user must consider whether in reality the new technology would likely be
applied to those vehicles which received the previous  technology or those which did not,
or a combination of the two. The effectiveness of adding the subsequent technology may
depend on which vehicles are receiving it.

                                       4-8

-------
                                    Results of Proposed and Alternative Standards

             In OMEGA, the costs and effectiveness of technologies are assumed to be
the same for all vehicle models that belong to the sale vehicle type category. There may
be cases when a vehicle model in the baseline may already contain some CO2-reducing
technology; OMEGA considers this when determining whether a technology can or
cannot be applied to it. In the inputs to the model, the user can limit the volume of a
specific vehicle model's sales which can receive a technology package by indicating the
fraction of its baseline that already contains some effectiveness and cost of each specific
technology package. In addition, as described above, the volume of a given vehicle
type's sales which can receive a specific technology package can also be limited in an
input file with a market penetration "cap", if desired.  The effectiveness and application
limits of each technology package can vary  over time, if desired.

             OMEGA adds technology effectiveness according to the following
equation in which the subscripts t and t-1 represent the times before and after technology
addition, respectively.  The numerator the effectiveness of the current technology
package and the denominator serves to "back out" any effectiveness that is present in the
baseline. CAP refers to the market penetration cap, AIE is the "average incremental
effectiveness" of the technology package on a vehicle type, and TEB is the "technology
effectiveness basis", which denotes the fraction of the technology present in the baseline.
                                         1-AIExTEB
             OMEGA then adds technology cost according to the equations below,
where CEB refers to the "cost effectiveness basis", or in other words, the technology cost
that is present in the baseline.
                        IncrementalCost = TechCost * (CAP - CEB)
                         ,, ,   , „          TechCost *ModelSales
                     AvgVehicleCost A
                                             TotalFleetSales
             EPA's OMEGA model calculates the new CO2 and average vehicle cost
after each technology package has been added. To simplify the model's algorithm, EPA
has chosen to input the package costs and effectiveness values on a step-wise basis.  This
is not the same "incremental" approach implemented in the Volpe model because each
step in OMEGA has incorporated several technologies. However, for simplification in

                                       4-9

-------
Draft Regulatory Impact Analysis

the core model calculations, the user must enter into the technology input file the
technology costs which are incremental to the technology package immediately preceding
it. In the case of the first technology package, this is simply the full technology package
cost, since it is going on a baseline vehicle and since any technology in the baseline is
considered in the equations, as described in the equations above.
4.3  Modeling Process

       In order to determine the technology costs associated with this proposed rule,
EPA performed two separate modeling exercises. The first was to determine the costs
associated with meeting any existing regulation of CO2 or MPG.  The latest regulation
that has been promulgated is NHTSA's CAFE program for MY 2011, directed under the
Energy Independence and Security Act (EISA). EPA considers the MY 2011  CAFE
regulations to constitute the "reference case" for calculating the costs and benefits of this
GHG proposal. In other words, absent any further rulemaking, this is the vehicle  fleet
EPA would expect to see through 2016; the "status quo".  In order to calculate the costs
and benefits of this proposed rule alone, EPA seeks to subtract out any costs associated
with meeting any existing standards related to GHG emissions. EPA ran OMEGA a
second time to calculate the cost of meeting the EPA's proposed standards in 2016, and
then subtracted the results of the reference case model run to determine the costs of this
proposed GHG program.

       Conceptually, OMEGA begins by determining the specific CO2 emission
standard applicable for each manufacturer and its vehicle class (i.e., car or truck).  Since
the proposed 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 proposed regulations which would 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 truck and their footprint values.
When evaluating the 2011 MY CAFE standards, the car-truck trading was limited to 1.2
mpg. When evaluating the proposed 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  proposed 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 DRIA to this proposed rule. When evaluating
the 2016 standards using OMEGA, the proposed CO2 standard which manufacturers
would 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  would 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.
                                      4-10

-------
                                    Results of Proposed and Alternative Standards
       The cost of the improved A/C systems required to generate the 11 g/mi credit was
estimated separately. This is consistent with the proposed A/C credit procedures, which
would 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 would likely apply them.
Conceptually, this approach estimates the cost of adding the technology from the
manufacturer's perspective and divides it by the mass of CC>2 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 purchaser's 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, and 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:
                                       4-11

-------
Draft Regulatory Impact Analysis
                                   pp                    i
                        TechCost - ^ [dFSt x VMTt ] >
       ManufCostEff =	—	^	,_\_Gap)_
       Where

       ManufCostEff = Manufacturer-Based Cost Effectiveness (in dollars per kilogram
C02),

       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,

       dFSj = Difference in fuel consumption due to the addition of technology times
fuel price in year i,

       dCCh = Difference  in CCh 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,

       1- Gap = Ratio of onroad fuel economy to two-cycle (FTP/HFET) fuel economy
       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 proposal, 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. It is reasonable to estimate that the added technology to improve CC>2
level and fuel economy would retain this same percentage of value when the vehicle is
five years old. However, it is less clear whether first purchasers, and thus, manufacturers
would consider this residual value when ranking technologies and making vehicle
purchases, respectively.  For this proposal, this factor was not included in the
determination of manufacturer-based net cost-effectiveness in the analyses performed in
                                       4-12

-------
                                    Results of Proposed and Alternative Standards

support of this proposed rule.  Comments are requested on the benefit of including an
increase in the vehicle's residual value after five years in the calculation of effective cost.

       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
CC>2 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-cyUnder 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 an incremental
reduction in fuel consumption depends on the CO2 level of the vehicle prior to adding the
technology.  EPA requests comment on the use of manufacturer-based net cost-
effectiveness to rank CO2 emission reduction technologies in the  context of evaluating
alternative fleet average standards for this rule. EPA believes this manufacturer-based
net cost-effectiveness metric is appropriate for ranking technology in this proposed
program because it considers effectiveness values that may vary widely among
technology packages when determining the order of technology addition.  Comments are
requested on this option and on any others thought to be appropriate.
4.4 Modeling of CAA Compliance Flexibilities

       EPA's proposed rule incorporates several compliance flexibilities. Three of these
flexibilities, the credit for air conditioning system improvements, car-truck credit trading,
and FFV credits, are expected to be used extensively by manufacturers and have been
factored into our estimates of the cost of the proposed CO2 standards.  OMEGA was
designed to be able to address the first two types of flexibilities directly through the
appropriate specification of model inputs and scenario definition. However, for several
reasons, the expected impact of A/C credits was handled outside of OMEGA. The
impact of car-truck credit trading was accomplished in a slightly more complex fashion
than will be the case with future versions of the model.  OMEGA was not originally
designed to include FFV credits in terms of miles  per gallon. The methods used to
account for these three flexibilities are described below.

       OMEGA is capable of including both the impact of air conditioning use on CO2
emissions from the tailpipe (indirect A/C emissions) and refrigerant emissions (direct
A/C emissions). The current approach to specifying refrigerant emissions in the Market
file and the effectiveness of refrigerant emission control in the Technology file allows for
the straightforward accounting of EPA's current approach to estimating both of these
factors. As described in Chapter 2 above, EPA currently estimates the same base level of
direct A/C emissions from cars and a distinct level of emissions from trucks.  These
levels can be input directly into Column AD of the Market file. The reduction in direct
A/C emissions associated with improved A/C systems can be input into Column U of the
Technology file.
                                       4-13

-------
Draft Regulatory Impact Analysis

       Accounting for indirect A/C emissions, consistent with our approach to estimating
these emissions in Chapter 2, however, is more difficult. In Chapter 2, we estimate a
single level of 14 g/mi CO2 from A/C usage and a potential reduction of 40% for a high
efficiency A/C design (maximum A/C credit of 5.7 g/mi CO2).  OMEGA currently
combines all sources of CO2  tailpipe emissions (i.e., those measured over the 2-cycle
compliance test and those from A/C usage). Adding 14 g/mi CO2 from A/C usage to the
base emission level of all vehicles could be easily accomplished. However, specifying a
consistent 40% reduction of this  incremental emission level would not be. The CO2
effectiveness of technologies  included in the Technology file applies to all sources of
CO2 emissions. Since the base 2-cycle CO2 emission level of vehicles varies, the
additional 14 g/mi of indirect A/C emissions would represent a different percentage of
total CO2 emissions of each vehicle.  A single effectiveness value for the benefit of high
efficiency A/C systems would therefore produce a slightly different CO2 emission
reduction for each vehicle.

       In addition, OMEGA  is currently designed to include both indirect and direct A/C
emissions in the accounting of emissions towards compliance with the specified
standards. This means that the 14 g/mi of indirect A/C emissions and 17-21 g/mi of
direct A/C emissions are included in the base level of vehicles' emissions. Their
remaining levels after the application of technology are considered when determining
whether a manufacturer is in  compliance with the specified standards.  However, this is
not consistent with the design of the proposed A/C credit system. Neither direct nor
indirect A/C emissions are included in the compliance determination towards the
proposed CO2 emission standards. Compliance is determined based on CO2 emissions
measured over the 2-cycle test procedure which does not include these A/C emissions.
Then, reductions in A/C emissions are essentially subtracted from the measured 2-cycle
CO2 emissions.

       With the current OMEGA model design, it was more straightforward to determine
the total A/C credit applicable to each manufacturer in 2016 and adjust their proposed
CO2 emission standards accordingly. Thus, the effective 2016 proposed car and truck
standards were increased by 10.2 g/mi and 11.5 g/mi, respectively.  OMEGA was then
run to determine the level of non-A/C technology needed to meet the proposed standards
after accounting for A/C credits.  After modeling, EPA then added a uniform AC cost of
$60 per vehicle to each manufacturer's per vehicle technology cost.

       With respect to car-truck trading, OMEGA was originally designed to allow the
trading of car-truck credits on a vehicle-g/mi CO2 basis. For example, if a manufacturer
over-complies with its applicable CO2 standard for cars by 3 g/mi and it sells 1,000,000
cars, it generates 3,000,000 vehicle-g/mi CO2 worth of credits.  If these credits are used
to compensate for under-compliance towards the truck CO2 standard and truck sales are
500,000, the manufacturer's truck CO2 emission level could be as much as 6 g/mi CO2
above the standard.  This is the credit trading approach used in EPA's Tier 2 light-duty
vehicle emission control program.

       As described in section III.B.4 of the Preamble to this proposed rule, EPA is
proposing to trade car and truck  credits on a lifetime CO2 emission basis. In the above

                                      4-14

-------
                                    Results of Proposed and Alternative Standards

example, cars are assumed to have a lifetime VMT of 152,000 miles. Therefore, the
value of the 3 g/mi over-compliance is multiplied by 1,000,000 vehicles and 190,971
miles and converted to metric tons, or 573,000 metric tons of CO2. If these credits are
used to compensate for under-compliance towards the truck CO2 standard and truck sales
are again 500,000, the manufacturer's truck CO2 emission level could only be as much as
5.2 g/mi CO2 above the standard, as the lifetime VMT of trucks is 221,199 miles.

       In order to simulate trading on a lifetime emission basis, we ran the OMEGA
model with a single vehicle fleet (i.e., as if cars and trucks were being regulated
identically).  We adjusted the base CO2 emissions of trucks (in g/mi) to account for their
higher lifetime VMT by multiplying them by 1.158 (the ratio of 221,199 to 190,971).
We also adjusted the proposed CO2 emission standard applicable to each manufacturer in
2016 to account for these higher truck emissions. Because each manufacturer has a
different car-truck sales split, the degree to which their standards were increased varied.
Thus, each manufacturer's standard was specified as a universal or flat standard in the
Scenario file (each manufacturer's compliance was evaluated separately).  This universal
standard was essentially determined by applying the proposed 2016 car standard to each
manufacturer's cars and 1.158 times the proposed 2016 truck standard to each
manufacturer's trucks. The OMEGA input and output files using the latest version of the
model, including the adjusted truck base CO2 levels and adjusted standards can be found
under "EPA OMEGA Model" in the docket to this rule.

       Under the proposal, FFV credits are only available through model year 2015.
Since we use the OMEGA model directly to evaluate technical feasibility and costs only
for the 2016 model year, FFV credits are not a factor. (FFV  credits use in earlier years is
accounted for in projecting the cost of technology for 2012-2015 below.) However, as
discussed above, some manufacturers' 2008 baseline fleets (adjusted for projected sales
in 2011) do not meet the 2011 CAFE standards which comprise the reference case for
this analysis.  FFV credits are available under this program and expected be used at the
maximum allowable level by Chrysler,  Ford and General Motors for both their cars and
trucks and by Nissan for their trucks. Under the current CAFE program, FFV credits are
limited to 1.2 mpg in 2011. Car-truck trading is also allowed under this program, up to
1.0 mpg in 2011. However, our reference case is a 2016 vehicle fleet complying with the
2011 CAFE standards. In 2016, the limit on FFV credits is reduced to 0.8 mpg and car-
truck trading is increased to 1.5 mpg. We use these latter levels here.

       Because fuel economy is the inverse of fuel consumption, a specified change in
fuel economy (e.g., either the limit on FFV credits or car-truck trading) represents a
varying change in fuel consumption (and CO2 emissions) depending on the initial level
of fuel economy. For example, for a manufacturer whose truck standard is 22.5 mpg, its
trucks could be as low as 21 mpg if the manufacturer generated sufficient credits from its
car fleet. These two fuel economy levels represent CO2 emission levels of 395 and 423
g/mi, respectively, assuming all the vehicles are fueled with gasoline, a difference of 28
g/mi CO2. If the manufacturer's truck standard is 24 mpg, its trucks could be as low as
23 mpg if the manufacturer generated sufficient credits from its car fleet. These two fuel
economy levels represent CO2 emission levels of 363 and 386 g/mi, respectively, a
difference of 23 g/mi CO2. In both cases, the difference in terms of mpg is 1.5.

                                      4-15

-------
Draft Regulatory Impact Analysis

However, the difference in terms of CO2 emissions decreases as the base fuel economy
increases.

       As mentioned above, the OMEGA model is not designed to incorporate a credit
which is specified in terms of mpg. Since the effect in terms of CO2 emissions changes
with the level of fuel economy standard and the footprint-based 2011 CAFE standards
impose a different effective fuel economy standard on each manufacturer, these credits
cannot be simulated in an OMEGA model run which represents the CO2 standard as a
footprint-based standard. Thus, as was the case with the proposed 2016 standards, we
determined the effective flat standard applicable to each manufacturer's sales in 2011
after accounting for expected use of FFV credits. The 2011 CAFE standards for cars and
trucks were converted to CO2 emissions assuming that all vehicles were fueled with
gasoline (i.e., 8887/mpg). Truck emissions were again increased by a factor of 1.158 and
the car and truck standards for each manufacturer were sales weighted after again
increasing the truck standard by this same factor. An exception was made for those
manufacturers which traditionally pay CAFEE fines in lieu of compliance. For these
manufacturers, we substituted the achieved fuel economy levels from NHTSA's Volpe
Model evaluations of the 2011 CAFEE standards for these manufacturers' CAFEE
standards.  These manufacturers were BMW, Daimler, Porsche, Tata and Volkswagen.
Also, several manufacturers were not run through the model, as their 2008 vehicles
already complied with both 2011 car and truck CAFE standards: Honda, Kia, Mazda, and
Toyota.

       This approach allows unlimited trading of car-truck credits to occur.  This is
consistent with the proposal for 2016, but not 2011. In order to determine whether the
1.5 mpg car-truck trading limit had been exceeded, EPA converted the final car and truck
CO2 emissions levels for each manufacturer to their fuel economy equivalents and
compared these values to the applicable 2011 CAFE standards.  If the fuel economy  of
the vehicle class which under-complied with its standard exceeded 1.5 mpg, the shortfall
was reduced to 1.5 mpg and the fuel economy of the over-complying class decreased.
The OMEGA model was then run for cars and trucks  separately to determine the overall
cost of these trading limited manufacturers. Only two manufacturers were found to
exceed the trading limit in the unlimited trading runs, Daimler and Hyundai. The impact
of limiting trading on projected technology costs in 2011 was very small, $8 and $3 per
vehicle, respectively.  Hyundai's 2008 vehicles complied with the 2011  CAFEE
standards with unlimited trading (i.e., zero compliance cost).  However, limited trading
required Hyundai to apply technology to their trucks.  The OMEGA input and  output
files using the latest version of the model, including the adjusted truck base CO2 levels
and adjusted standards can be found under "EPA OMEGA Model" in the docket to this
rule.
4.5 Per Vehicle Costs 2012-2016

       As described above, the per-vehicle technology costs for this program alone must
account for any cost that incurred by compliance with existing vehicle programs. EPA

                                      4-16

-------
                                      Results of Proposed and Alternative Standards

first used OMEGA to calculate costs reflected in the existing CAFE program.  OMEGA
estimates that, on average, manufacturers will need to spend $78 per vehicle to meet the
current MY 2011 CAFE standards. A  Reference case costs are provided in Table 4-5
below.

                 Table 4-5 Incremental Technology Cost of the Reference Case

BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Cars
$ 319
$ 7
$ 431
$ 28
$ 28
$ 0
$ 0
$ 0
$ 0
$ 96
$ 0
$ 535
$ 64
$ 99
$ 691
$ 0
$ 269
$ 47
Trucks
$ 479
$ 125
$ 632
$ 211
$ 136
0
$ 76
$ 48
$ 0
$ 322
$ 19
$ 1,074
$ 100
$ 231
$ 1,574
$ 0
$ 758
$ 141
Combined
$ 361
$ 59
$ 495
$ 109
$ 73
$ 0
$ 14
$ 8
$ 0
$ 123
$ 6
$ 706
$ 77
$ 133
$ 1,161
$ 0
$ 354
$ 78
       EPA then used OMEGA to calculate the costs of meeting the proposed 2016
standards, which are displayed in Table 4-6 below, and two alternative scenarios for
sensitivity.  In Table 4-13 and Table 4-14, EPA presents the per-vehicle cost for these
scenarios, respectively.  EPA has accounted for the cost to meet the standards in the
reference case. In other words, the following tables contain results of the OMEGA
control case runs after the reference case values have been subtracted.
A It should be noted that the latest version of OMEGA projects slightly different costs than those shown
here. This is usually due to an error when the model eliminates over-compliance which occurs with the last
step of technology addition. The costs presented here reflect the correction of this error. The latest version
of the model also reflects several improvements to the model's algorithms when selecting between car and
truck control. These revisions generally only change the projected cost by a dollar or two per vehicle and
do not affect the overall conclusions of this analysis.
                                         4-17

-------
Draft Regulatory Impact Analysis
          Table 4-6 Incremental Technology Cost of the Proposed 2016 CO2 Standards

BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Cars
$1,701
$1,331
$1,631
$1,435
$969
$606
$739
$741
$946
$1,067
$1,013
$1,549
$903
$1,093
$1,270
$600
$1,626
$968
Trucks
$1,665
$1,505
$1,357
$1,485
$1,782
$695
$1,680
$1,177
$1,030
$1,263
$1,194
$666
$1,329
$1,263
$674
$436
$949
$1,214
All
$1,691
$1,408
$1,543
$1,457
$1,311
$633
$907
$812
$958
$1,090
$1,064
$1,268
$1,057
$1,137
$952
$546
$1,509
$1,051
       EPA estimates that the additional technology required for manufacturers to meet
the GHG standards in this proposed rule will cost on average $105I/vehicle.

       The majority of manufacturers representing the vast majority of sales in 2016 are
projected to comply with the proposed 2016 standards with the addition of technology
under the penetration limits described above.  However, several smaller volume
manufacturers (at least with respect to U.S. sales) are projected to fall short of
compliance. The CO2 standards applicable to each manufacturer based on its distribution
of sales and vehicle footprints are shown in Table 4-7 along with the projected achieved
level of CO2 emissions from the OMEGA model.
                                       4-18

-------
                                   Results of Proposed and Alternative Standards
              Table 4-7 CO2 Standards and Projected Achieved Levels in 2016

Manufacturer
BMW*
Chrysler
Daimler *
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche *
Subaru
Suzuki
Tata*
Toyota
Volkswagen *
Overall
Car
Standard
237.1
241.7
244.1
238.5
238.5
231.6
231.8
233.7
230.5
227.5
235.3
214.8
224.7
216.8
258.6
230
227.2
234
Achieved CO2
245.8
220.1
246
225.1
221.2
186.3
196.8
191.6
205.6
212.4
212.5
247.8
208.3
197.4
270.6
188.1
240.6
208.8
Truck
Standard
295.2
312.2
305.6
326
333.9
292.7
290.4
300.7
283.4
281.7
315.7
298.7
279.1
283.7
287.1
303.2
304.8
314
Achieved CO2
339.7
334.1
363
353.6
349.4
302.8
308
310.5
285.4
319.7
351.5
398.5
270.7
312.1
397.4
307.3
394.9
334.6
4.6 Technology Penetration

         The major technologies chosen by OMEGA are described in the Table 4-8
through Table 4-11 below for the reference case and control case cars and trucks. The
values in the table containing the control case technology are for that alone - EPA has
subtracted out the impact of the reference case.
                                      4-19

-------
Regulatory Impact Analysis
                                           Table 4-8 2016 Technology Penetration in the Reference Case-Cars
Cars
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
85.9%
48.0%
47.0%
12.7%
54.9%
0.0%
29.9%
24.2%
96.0%
77.8%
99.8%
31.1%
48.7%
100.0%
64.4%
99.7%
16.8%
53.0%
VVTL
11.9%
0.0%
0.0%
0.0%
0.0%
100.0%
0.0%
0.0%
3.3%
0.0%
0.0%
0.0%
3.3%
0.0%
0.0%
0.0%
0.7%
15.0%
GDI
8.9%
0.0%
5.2%
0.0%
5.6%
0.0%
0.0%
0.0%
13.4%
0.0%
5.1%
0.0%
0.0%
0.0%
31.1%
7.1%
41.0%
5.6%
GDI
Deac
9.0%
0.0%
15.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
14.8%
0.0%
0.0%
33.4%
0.0%
4.7%
1.0%
GDI
Turbo
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
16.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.1%
Diesel
0.0%
0.0%
1.6%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
6Spd
Auto
88.5%
23.0%
45.1%
36.8%
13.2%
5.9%
0.1%
0.8%
36.0%
77.4%
26.6%
0.0%
26.9%
100.0%
31.1%
35.7%
73.6%
27.8%
Wet
DCT
9.0%
0.0%
31.9%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
31.1%
0.0%
0.0%
33.4%
0.0%
17.9%
2.0%
Dry
DCT
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
42 S-
s
9.0%
0.0%
15.2%
0.0%
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
14.8%
0.0%
0.0%
33.4%
0.0%
4.7%
1.0%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
3.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.4%
Power
Split
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
15.8%
0.0%
3.1%
2-
Mode
0.0%
0.0%
0.0%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
EPS
0.0%
2.9%
0.0%
3.6%
0.0%
0.0%
0.0%
2.5%
0.6%
4.9%
1.9%
0.0%
0.0%
0.0%
0.0%
63.8%
0.0%
13.3%



















                                                                    4-20

-------
                                                   Results of Proposed and Alternative Standards
Table 4-9 2016 Technology Penetration in the Reference Case-Trucks
Trucks
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
97.0%
4.8%
92.9%
37.7%
24.6%
0.0%
28.6%
98.8%
49.9%
100.0%
100.0%
91.6%
32.6%
100.0%
85.0%
85.5%
99.3%
48.6%
VVTL
0.0%
0.0%
0.0%
0.0%
0.0%
100.0%
0.0%
0.0%
0.0%
0.0%
49.5%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
16.0%
GDI
3.0%
0.1%
40.3%
1.3%
0.0%
4.2%
0.0%
0.0%
2.5%
0.0%
49.5%
0.0%
0.0%
17.8%
0.0%
8.1%
100.0%
9.0%
GDI
Deac
16.9%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
18.4%
0.0%
47.0%
0.0%
0.0%
85.0%
0.0%
17.3%
1.6%
GDI
Turbo
0.0%
0.0%
44.7%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
38.0%
0.0%
0.0%
0.0%
0.0%
67.7%
2.2%
Diesel
0.0%
0.1%
16.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.7%
0.3%
6Spd
Auto
80.1%
34.1%
47.3%
17.2%
14.3%
0.0%
27.0%
1.2%
43.5%
64.9%
50.1%
0.0%
32.6%
100.0%
0.0%
20.0%
0.0%
20.5%
Wet
DCT
16.9%
0.0%
44.7%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
18.4%
0.0%
85.0%
0.0%
0.0%
85.0%
0.0%
85.0%
3.8%
Dry
DCT
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
42 S-
s
16.9%
0.0%
44.7%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
18.4%
0.0%
85.0%
0.0%
0.0%
85.0%
0.0%
85.0%
3.8%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Power
Split
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
6.6%
0.0%
1.3%
2-
Mode
0.0%
0.0%
0.0%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
EPS
0.0%
2.9%
0.0%
3.6%
0.0%
0.0%
0.0%
2.5%
0.6%
4.9%
1.9%
0.0%
0.0%
0.0%
0.0%
63.8%
0.0%
13.3%



















                              4-21

-------
Regulatory Impact Analysis
                                           Table 4-10 2016 Technology Penetration in the Control Case-Cars
Cars
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
97.8%
46.2%
89.2%
51.9%
43.5%
0.0%
16.9%
4.5%
64.3%
21.8%
63.8%
100.0%
14.6%
0.0%
100.0%
64.9%
87.4%
47.0%
VVTL
2.0%
44.3%
0.0%
27.0%
37.2%
100.0%
31.3%
35.6%
28.4%
65.5%
36.2%
0.0%
42.5%
76.2%
0.0%
34.8%
0.1%
40.3%
GDI
4.9%
59.0%
3.4%
27.0%
40.6%
24.0%
34.1%
35.6%
61.3%
74.1%
36.2%
2.6%
42.4%
76.2%
3.6%
41.6%
7.7%
33.2%
GDI
Deac
37.2%
19.1%
42.0%
37.7%
11.0%
0.0%
1.8%
0.0%
1.5%
0.0%
22.6%
27.2%
3.5%
0.0%
81.4%
2.3%
25.2%
13.8%
GDI
Turbo
44.1%
0.8%
39.9%
12.2%
2.9%
0.0%
0.8%
0.0%
11.6%
0.0%
3.2%
57.8%
6.6%
0.0%
0.0%
0.0%
58.4%
8.0%
Diesel
0.0%
0.0%
1.6%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
6Spd
Auto
14.8%
24.2%
9.4%
11.1%
5.0%
5.8%
2.8%
0.0%
34.5%
22.6%
33.3%
7.8%
0.4%
0.0%
13.7%
31.2%
11.9%
15.6%
Wet
DCT
65.5%
18.0%
60.8%
46.0%
13.4%
0.0%
2.6%
0.0%
13.1%
0.0%
25.8%
70.0%
10.1%
0.0%
70.0%
26.5%
72.0%
24.3%
Dry
DCT
5.7%
44.3%
14.2%
27.0%
37.2%
24.0%
31.3%
35.6%
28.2%
65.5%
36.2%
0.0%
42.4%
76.2%
0.0%
21.1%
0.0%
27.1%
42 S-
S
71.0%
62.2%
72.5%
73.0%
50.6%
24.0%
33.9%
35.6%
41.3%
65.5%
60.5%
70.0%
52.6%
76.2%
70.0%
23.4%
70.0%
46.4%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
3.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.4%
Power
Split
3.9%
0.0%
3.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.0%
0.0%
0.0%
0.0%
3.6%
15.8%
1.4%
3.4%
2-
Mode
10.3%
0.0%
9.3%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
15.0%
0.0%
0.0%
11.4%
0.0%
13.5%
1.2%
EPS
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
R


















                                                                    4-22

-------
                                                   Results of Proposed and Alternative Standards
Table 4-11 2016 Technology Penetration in the Control Case-Trucks
Trucks
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
99.5%
81.6%
92.9%
84.5%
83.1%
30.7%
89.3%
77.3%
92.9%
90.3%
98.6%
85.0%
34.4%
85.0%
100.0%
93.6%
87.9%
80.7%
VVTL
0.0%
4.1%
0.0%
6.1%
3.5%
69.3%
0.0%
0.0%
7.0%
0.0%
1.4%
15.0%
57.0%
0.0%
0.0%
0.0%
11.9%
11.8%
GDI
0.5%
41.1%
2.4%
32.3%
24.3%
24.4%
0.0%
42.7%
9.9%
0.0%
11.8%
15.0%
51.5%
35.7%
4.1%
27.4%
15.0%
24.5%
GDI
Deac
29.0%
39.3%
47.8%
39.6%
45.6%
3.5%
8.5%
0.0%
5.8%
18.4%
36.0%
53.7%
5.8%
17.8%
80.9%
2.1%
29.2%
27.3%
GDI
Turbo
56.0%
4.6%
37.2%
13.2%
13.0%
7.9%
76.5%
33.0%
40.8%
55.2%
8.4%
31.3%
27.7%
31.6%
0.0%
0.0%
55.8%
14.0%
Diesel
0.0%
0.1%
16.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.7%
0.3%
6Spd
Auto
15.0%
52.5%
15.0%
28.6%
23.0%
19.3%
4.0%
42.7%
7.2%
16.7%
34.1%
15.0%
0.0%
35.7%
15.0%
28.2%
15.0%
25.0%
Wet
DCT
70.0%
33.5%
70.1%
53.2%
58.6%
11.5%
85.0%
33.0%
46.6%
73.6%
44.4%
70.0%
69.3%
49.3%
70.0%
2.1%
70.0%
40.6%
Dry
DCT
0.0%
4.1%
0.0%
5.7%
3.5%
4.5%
0.0%
0.0%
7.0%
0.0%
1.4%
0.0%
15.7%
0.0%
0.0%
0.0%
0.0%
3.1%
42 S-
S
70.0%
37.6%
70.1%
58.6%
62.1%
16.0%
85.0%
33.0%
53.6%
73.6%
45.9%
70.0%
49.2%
49.3%
70.0%
2.1%
70.0%
43.1%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Power
Split
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
4.1%
6.6%
0.0%
1.3%
2-
Mode
15.0%
0.0%
14.9%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
15.0%
0.0%
0.0%
10.9%
0.0%
15.0%
1.1%
EPS
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
-\
Rt


















                             4-23

-------
Regulatory Impact Analysis
       As can be seen, the overall reduction in vehicle weight is projected to be 4%. This
reduction varies across the two vehicle classes and vehicle base weight. For cars below 2950
pounds curb weight, the reduction is 2.3% (62 pounds), while it was 4.4% (154 pounds) for cars
above 2950 curb weight. For trucks below 3850 pounds curb weight, the reduction is 3.5% (119
pounds), while it was 4.5% (215 pounds) for trucks above 3850 curb weight.  Splitting trucks at a
higher weight, for trucks below 5000 pounds curb weight, the reduction is 3.3% (140 pounds),
while it was 6.7% (352 pounds) for trucks above 5000 curb weight. These results are tabulated
below in Table 4-12.

                  Table 4-12 Breakdown of Weight Reduction in Modeling Results

Cars
Trucks with 3850
Ib break point
Trucks with 5000
Ib break point
Weight
Category
< 2950 Ibs
> 2950 Ibs
< 3850 Ibs
> 3850 Ibs
< 5000 Ibs
> 5000 Ibs
% Weight
Reduction
2.3%
4.4%
3.5%
4.5%
3.3%
6.7%
Average
Weight
Reduction
62 Ibs
154 Ibs
119 Ibs
21 5 Ibs
140 Ibs
352 Ibs
4.7 Manufacturer-Specific Standards

       As described in Section 3.2, in any attribute-based regulatory structure, manufacturers
are bound to have different overall GHG targets, since they are based on the size and sales mix
of each manufacturer.  The fleet-wide averages calculated for the proposed 2016 model year are
presented in Table 4-13.
                                         4-24

-------
                                          Results of Proposed and Alternative Standards
                      Table 4-13 2016 Projected Standards by Manufacturer


BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Standards
Cars
226.9
231.5
233.9
228.3
228.3
221.4
221.6
223.5
220.3
217.3
225.1
204.6
214.5
206.6
248.4
219.8
217.0
223.8
Trucks
283.7
300.7
294.1
314.5
322.4
281.2
278.9
289.2
271.9
270.2
304.2
287.2
267.6
272.2
275.6
291.7
293.3
302.5
Combined
241.8
262.1
253.2
266.5
267.9
239.2
231.8
234.2
227.7
223.5
247.5
230.8
233.7
223.4
262.8
243.5
230.3
250.2
       These car and truck standards average out to an overall industry CO2 stringency of 250
g/mi, consistent with President Obama's announcement on May 19, 2009.

4.8 Alternative Program Stringencies

        EPA analyzed the technology cost of two alternative stringency scenarios: 4%/year and
6%/year. With the reference case the same as that described above in Section 4.1, the costs of
the two alternative control cases along with the technology penetrations and the manufacturer-
specific standards are presented in Table 4-14 through Table 4-19, below.  The manufacturers's
CO2 targets for these alternative standards are presented in Table 4-20 and Table 4-21.
                                             4-25

-------
Regulatory Impact Analysis
                     Table 4-14 2016 Technology Cost in the 4% sensitivity case

BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Cars
$ 1,701
$ 1,167
$ 1,631
$ 1,339
$ 850
$ 606
$ 685
$ 741
$ 819
$ 1,004
$ 910
$ 1,549
$ 903
$ 1,093
$ 1,270
$ 479
$ 1,626
$ 893
Trucks
$ 1,665
$ 1,505
$ 1,357
$ 1,417
$ 1,769
$ 460
$ 1,505
$ 738
$ 1,030
$ 1,263
$ 1,194
$ 666
$ 1,131
$ 1,026
$ 674
$ 436
$ 949
$ 1,154
All
$ 1,691
$ 1,316
$ 1,543
$ 1,373
$ 1,237
$ 563
$ 832
$ 741
$ 849
$ 1,034
$ 991
$ 1,268
$ 985
$ 1,076
$ 952
$ 465
$ 1,509
$ 980
                                           4-26

-------
                       Results of Proposed and Alternative Standards
Table 4-15 2016 Technology Cost in the 6% sensitivity case

BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Cars
$ 1,701
$ 1,642
$ 1,631
$ 2,175
$ 1,718
$ 777
$ 1,275
$ 1,104
$ 1,321
$ 1,495
$ 1,654
$ 1,549
$ 1,440
$ 1,718
$ 1,270
$ 755
$ 1,626
$ 1,381
Trucks
$ 1,665
$ 2,211
$ 1,357
$ 2,396
$ 2,158
$ 1,580
$ 1,680
$ 1,772
$ 1,293
$ 2,065
$ 2,274
$ 666
$ 1,615
$ 2,219
$ 674
$ 1,182
$ 949
$ 1,866
All
$ 1,691
$ 1,893
$ 1,543
$ 2,273
$ 1,903
$ 1,016
$ 1,347
$ 1,213
$ 1,317
$ 1,563
$ 1,830
$ 1,268
$ 1,503
$ 1,846
$ 952
$ 895
$ 1,509
$ 1,544
                          4-27

-------
Regulatory Impact Analysis
                                         Table 4-16 2016 Technology Penetration in the 4% sensitivity case- Cars
Cars
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
97.8%
33.3%
89.2%
50.0%
56.1%
0.0%
17.4%
4.5%
40.7%
29.2%
63.8%
100.0%
14.6%
0.0%
100.0%
82.0%
87.4%
51.3%
VVTL
2.0%
44.3%
0.0%
27.0%
23.8%
100.0%
31.3%
35.6%
28.4%
58.2%
36.2%
0.0%
42.5%
76.2%
0.0%
17.7%
0.1%
34.8%
GDI
4.9%
51.7%
3.4%
36.1%
40.3%
24.0%
31.3%
35.6%
37.7%
74.1%
42.3%
2.6%
42.4%
76.2%
3.6%
24.5%
7.7%
30.7%
GDI
Deac
37.2%
13.6%
42.0%
26.8%
10.6%
0.0%
0.0%
0.0%
1.5%
0.0%
16.6%
27.2%
3.5%
0.0%
81.4%
2.3%
25.2%
11.4%
GDI
Turbo
44.1%
0.8%
39.9%
12.2%
2.9%
0.0%
0.8%
0.0%
11.6%
0.0%
0.0%
57.8%
6.6%
0.0%
0.0%
0.0%
58.4%
7.7%
Diesel
0.0%
0.0%
1.6%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
6Spd
Auto
14.8%
16.9%
9.4%
16.2%
17.6%
5.8%
0.0%
0.0%
10.9%
29.9%
41.9%
7.8%
0.4%
0.0%
13.7%
31.2%
11.9%
18.3%
Wet
DCT
65.5%
12.4%
60.8%
39.0%
13.4%
0.0%
0.8%
0.0%
13.1%
0.0%
16.6%
70.0%
10.1%
0.0%
70.0%
43.6%
72.0%
25.6%
Dry
DCT
5.7%
44.3%
14.2%
27.0%
23.8%
24.0%
31.3%
35.6%
28.2%
58.2%
36.2%
0.0%
42.4%
76.2%
0.0%
4.0%
0.0%
21.6%
42 S-
S
71.0%
56.7%
72.5%
66.1%
37.3%
24.0%
32.1%
35.6%
41.3%
58.2%
52.8%
70.0%
52.6%
76.2%
70.0%
6.3%
70.0%
39.0%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
3.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.4%
Power
Split
3.9%
0.0%
3.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.0%
0.0%
0.0%
0.0%
3.6%
15.8%
1.4%
3.4%
2-
Mode
10.3%
0.0%
9.3%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
15.0%
0.0%
0.0%
11.4%
0.0%
13.5%
1.2%
EPS
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
I


















                                                                      4-28

-------
                                                      Results of Proposed and Alternative Standards
Table 4-17 2016 Technology Penetration in the 4% sensitivity case- Trucks
Trucks
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
99.5%
81.6%
92.9%
84.5%
82.0%
10.3%
84.0%
34.6%
92.9%
90.3%
98.6%
85.0%
61.5%
85.0%
100.0%
93.6%
87.9%
77.6%
VVTL
0.0%
4.1%
0.0%
2.8%
3.5%
69.3%
0.0%
0.0%
7.0%
0.0%
1.4%
15.0%
29.9%
0.0%
0.0%
0.0%
11.9%
10.7%
GDI
0.5%
41.1%
2.4%
29.9%
23.1%
11.9%
7.8%
7.7%
9.9%
0.0%
11.8%
15.0%
51.5%
59.5%
4.1%
27.4%
15.0%
22.1%
GDI
Deac
29.0%
39.3%
47.8%
38.7%
45.6%
3.5%
8.5%
0.0%
5.8%
18.4%
36.0%
53.7%
5.8%
17.8%
80.9%
2.1%
29.2%
27.1%
GDI
Turbo
56.0%
4.6%
37.2%
13.2%
13.0%
0.0%
63.5%
25.3%
40.8%
55.2%
8.4%
31.3%
27.7%
7.7%
0.0%
0.0%
55.8%
12.6%
Diesel
0.0%
0.1%
16.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.7%
0.3%
6Spd
Auto
15.0%
52.5%
15.0%
31.1%
21.8%
6.8%
11.8%
7.7%
7.2%
16.7%
34.1%
15.0%
27.1%
59.5%
15.0%
28.2%
15.0%
24.0%
Wet
DCT
70.0%
33.5%
70.1%
51.9%
58.6%
3.5%
71.9%
25.3%
46.6%
73.6%
44.4%
70.0%
57.9%
25.5%
70.0%
2.1%
70.0%
38.7%
Dry
DCT
0.0%
4.1%
0.0%
2.8%
3.5%
4.5%
0.0%
0.0%
7.0%
0.0%
1.4%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
2.2%
42 S-
S
70.0%
37.6%
70.1%
54.7%
62.1%
8.1%
71.9%
25.3%
53.6%
73.6%
45.9%
70.0%
33.5%
25.5%
70.0%
2.1%
70.0%
40.6%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Power
Split
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
4.1%
6.6%
0.0%
1.3%
2 -Mode
15.0%
0.0%
14.9%
0.0%
0.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
15.0%
0.0%
0.0%
10.9%
0.0%
15.0%
1.1%
EPS
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%



















                                4-29

-------
Regulatory Impact Analysis
                                         Table 4-18 2016 Technology Penetration in the 6% Sensitivity Case-Cars
Cars
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
97.8%
43.7%
89.2%
79.3%
53.8%
5.7%
39.8%
35.2%
32.5%
28.1%
61.8%
100.0%
14.6%
85.0%
100.0%
54.4%
87.4%
52.8%
VVTL
2.0%
48.5%
0.0%
7.6%
39.4%
86.5%
35.2%
35.6%
67.3%
64.2%
38.2%
0.0%
75.0%
0.0%
0.0%
45.3%
0.1%
39.7%
GDI
4.9%
48.5%
3.4%
13.3%
40.2%
41.1%
43.4%
56.8%
68.5%
66.8%
46.5%
2.6%
74.9%
1.5%
3.6%
52.2%
7.7%
37.9%
GDI
Deac
37.2%
33.6%
42.0%
35.8%
42.7%
0.0%
9.3%
0.0%
1.5%
4.3%
19.3%
27.2%
3.5%
7.2%
81.4%
2.3%
25.2%
19.1%
GDI
Turbo
44.1%
2.9%
39.9%
35.8%
2.9%
0.0%
17.8%
13.2%
11.6%
13.9%
19.2%
57.8%
6.6%
76.2%
0.0%
0.0%
58.4%
14.8%
Diesel
0.0%
0.0%
1.6%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
6Spd
Auto
14.8%
3.7%
9.4%
5.5%
8.1%
11.0%
8.2%
21.3%
2.7%
6.9%
10.8%
7.8%
0.4%
0.0%
13.7%
31.2%
11.9%
13.4%
Wet
DCT
65.5%
36.2%
60.8%
41.1%
39.5%
0.0%
27.1%
13.2%
13.1%
7.1%
21.7%
70.0%
17.6%
7.2%
70.0%
2.3%
72.0%
24.9%
Dry
DCT
5.7%
48.5%
14.2%
32.0%
39.4%
35.4%
35.2%
35.6%
67.2%
74.1%
53.6%
0.0%
67.4%
76.2%
0.0%
45.3%
0.0%
38.1%
42 S-
S
71.0%
84.7%
72.5%
73.1%
78.9%
35.4%
62.3%
48.7%
80.3%
81.2%
75.3%
70.0%
77.6%
83.5%
70.0%
47.6%
70.0%
62.8%
IMA
0.0%
0.0%
0.0%
1.6%
0.0%
3.0%
0.0%
0.0%
0.0%
1.9%
4.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.1%
Power
Split
3.9%
0.0%
3.2%
5.5%
0.0%
0.0%
0.0%
0.0%
0.0%
1.2%
5.7%
0.0%
0.0%
1.5%
3.6%
15.8%
1.4%
4.7%
2-
Mode
10.3%
0.0%
9.3%
4.8%
0.1%
0.0%
0.0%
0.0%
0.0%
0.8%
0.7%
15.0%
0.0%
0.0%
11.4%
0.0%
13.5%
2.0%
EPS
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
We
Redi


















                                                                     4-30

-------
                                                      Results of Proposed and Alternative Standards
Table 4-19 2016 Technology Penetration in the 6% Sensitivity Case-Trucks
Trucks
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
VVT
99.5%
81.6%
92.9%
90.7%
87.0%
54.0%
89.3%
85.4%
63.4%
90.3%
100.0%
85.0%
34.4%
85.0%
100.0%
82.0%
87.9%
83.4%
VVTL
0.0%
4.1%
0.0%
0.0%
1.7%
46.0%
0.0%
1.1%
36.5%
0.0%
0.0%
15.0%
57.0%
0.0%
0.0%
3.2%
11.9%
8.1%
GDI
0.5%
4.1%
2.4%
2.2%
1.7%
31.6%
0.0%
1.1%
41.3%
3.2%
0.8%
15.0%
51.5%
3.1%
4.1%
15.0%
15.0%
9.2%
GDI
Deac
29.0%
70.1%
47.8%
38.0%
68.5%
3.5%
8.5%
0.0%
5.8%
24.9%
9.8%
53.7%
5.8%
26.5%
80.9%
17.6%
29.2%
33.8%
GDI
Turbo
56.0%
10.8%
37.2%
45.0%
14.8%
50.4%
76.5%
83.9%
40.8%
56.9%
74.4%
31.3%
27.7%
55.4%
0.0%
34.2%
55.8%
39.0%
Diesel
0.0%
0.1%
16.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.7%
0.3%
6Spd
Auto
15.0%
5.1%
15.0%
2.6%
2.1%
0.0%
4.0%
0.0%
9.0%
5.3%
3.7%
15.0%
0.0%
0.0%
15.0%
13.6%
15.0%
5.3%
Wet
DCT
70.0%
80.9%
70.1%
71.0%
81.5%
54.0%
85.0%
83.9%
76.2%
60.6%
59.2%
70.0%
33.5%
70.0%
70.0%
51.8%
70.0%
67.0%
Dry
DCT
0.0%
4.1%
0.0%
5.0%
3.5%
31.0%
0.0%
1.1%
7.0%
9.4%
17.7%
0.0%
51.5%
0.0%
0.0%
3.2%
0.0%
8.5%
42 S-
S
70.0%
85.0%
70.1%
76.1%
85.0%
85.0%
85.0%
85.0%
53.6%
70.0%
76.9%
70.0%
85.0%
70.0%
70.0%
55.0%
70.0%
75.2%
IMA
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Power
Split
0.0%
0.0%
0.0%
3.1%
0.0%
0.0%
0.0%
0.0%
0.0%
5.3%
4.6%
0.0%
0.0%
3.1%
4.1%
6.6%
0.0%
2.3%
2-
Mode
15.0%
0.0%
14.9%
5.8%
0.1%
0.0%
0.0%
0.0%
0.0%
9.7%
3.6%
15.0%
0.0%
11.9%
10.9%
0.0%
15.0%
2.7%
EPS
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
We
Re*


















                                4-31

-------
Regulatory Impact Analysis
               Table 4-20 2016 Standards by Manufacturer in the 4% Sensitivity Case


BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Standards
Cars
230.3
235.0
237.4
231.7
231.8
224.8
225.0
226.9
223.7
220.7
228.5
208.0
217.9
210.0
251.8
223.2
220.5
227.2
Trucks
288.3
305.3
298.7
319.0
327.0
285.8
283.5
293.8
276.5
274.7
308.7
291.8
272.2
276.8
280.1
296.3
297.8
307.1
Combined
245.5
266.0
257.0
270.4
271.8
242.9
235.4
237.8
231.3
227.1
251.3
234.6
237.5
227.1
266.9
247.3
233.9
254.0
                                           4-32

-------
                                         Results of Proposed and Alternative Standards
              Table 4-21 2016 Standards by Manufacturer in the 6% Sensitivity Case


BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Standards
Cars
208.7
213.4
215.8
210.2
210.2
203.2
203.4
205.4
202.1
199.1
207.0
186.4
196.3
188.5
230.2
201.7
198.9
205.7
Trucks
259.4
276.4
269.8
290.1
298.1
256.9
254.6
264.9
247.6
245.8
279.8
262.9
243.3
247.9
251.2
267.4
268.9
278.2
Combined
222.0
241.2
233.1
245.6
247.2
219.2
212.6
215.1
208.7
204.7
227.6
210.7
213.3
203.6
241.4
223.3
211.1
230.0
4.9 Assessment of Manufacturer Differences

       The levels of requisite technologies shown above differ significantly across the
various manufacturers. This is to be expected for universal, or flat fuel economy or CO2
standards, since manufacturers' sales mixes differ dramatically in average size. However, use
of footprint-based standards should eliminate the effect of vehicle size, and thus, market mix,
on the relative stringency of a standard across manufacturers. Yet, large differences remain in
the level of technology projected to be required for various manufacturers to meet the
proposed standards. Therefore, several  analyses were performed to ascertain the cause of
these differences. Because the baseline case fleet consists of 2008 MY vehicle designs, these
analyses were focused on these vehicles, their technology and their CCh emission levels.

       Manufacturers' average CCh emissions vary for a wide range of reasons. In addition
to widely varying vehicle styles, designs, and sizes, manufacturers have implemented fuel
efficient technologies to varying degrees, as indicated in Table 4-22 below.
                                            4-33

-------
Regulatory Impact Analysis
       Table 4-22 Penetration of Technology in 2008 Vehicles with 2016 Sales: Cars and Trucks

BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
GDI
6.70%
0.00%
6.20%
0.60%
3.30%
1.20%
0.00%
0.00%
11.80%
0.00%
17.70%
0.00%
0.00%
0.00%
0.00%
7.50%
52.20%
6.40%
GDI+
Deac
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
GDI+
Turbo
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
Diesel
0.00%
0.00%
6.20%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.10%
0.10%
6 Speed
orCV
Trans
98.80%
27.90%
74.70%
28.10%
13.70%
4.20%
4.90%
0.90%
37.10%
76.10%
33.30%
3.90%
29.00%
100.00%
0.00%
30.60%
82.80%
27.10%
Dual
Clutch
Trans
0.80%
0.00%
11.40%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
10.90%
0.60%
Start-
Stop
0.00%
0.00%
0.00%
0.00%
0.10%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
Hybrid
0.10%
0.00%
0.00%
0.00%
0.10%
2.10%
0.00%
0.00%
0.00%
0.10%
0.00%
0.00%
0.00%
0.00%
0.00%
12.80%
0.00%
2.80%
       Once significant levels of technology are added to these vehicles in order to comply
with future standards, the impact of existing technology diminishes dramatically.
Manufacturers which did not utilize much technology in 2008 essentially catch up to those
which did.  The exception is the use of hybrid technology in 2008, since hybrids are not
projected to be needed by most manufacturers to meet the proposed standards. This primarily
affects Toyota, and to a lesser extent, Honda. Their use of hybrid technology in their 2008
fleet will continue to provide relatively greater CO2 reductions even in the 2016 projections.
As long as the vehicle designs of various manufacturers would produce the same level of CO2
emissions if their CO2 reducing technology was removed, for the most part, difference in the
application of technology in 2008 will not affect the level of technology needed in 2016.

       In addition, as mentioned above, differences in CO2 emissions due to differences the
distribution of sales by vehicle size should be largely eliminated by the use of a footprint-
based standard. Thus, just because a manufacturer produces larger vehicles than another
manufacturer does not explain the differences in required technology seen above.

       In order to focus this analysis on the 2008 MY fleet, it would be helpful to remove the
effect of differences in vehicle size and the use of CO2 reducing technology, so that the other
                                         4-34

-------
                                        Results of Proposed and Alternative Standards
causes of differences can be highlighted.  EPA used the EPA lumped parameter model
described in Chapter 1 to estimate the degree to which technology present on each 2008 MY
vehicle was improving fuel efficiency. The effect of this technology was then removed from
each vehicle to produce CO2 emissions which did not reflect any differences due to the use of
CO2 reducing technology.  This set of adjusted CO2 emission levels is referred to as "no
technology" emissions.

       The differences in the relative sizes of vehicles sold by each manufacturer were
accounted for by determining the difference between the sales-weighted average of each
manufacturer's "no technology" CCh levels and their required CCh emission level under the
proposed 2016 standards. This difference is the total reduction in CO2 emissions required for
each manufacturer relative to a "no technology" baseline. The same difference for the
industry as a whole is 71 g/mi CO2 for cars and 1.7 g/mi CO2 for trucks. This industry-wide
difference was subtracted from each manufacturer's difference to highlight which
manufacturers had lower and higher CO2 emission reduction requirements.  The results are
shown in Figure 4-1.
         Figure 4-1 CO2 Emissions Relative to Fleet Adjusted for Technology and Footprint

       As can be seen, the manufacturers projected in Table 4-22 to require the greatest levels
of technology also show the highest offsets relative to the industry. The greatest offset shown
in Figure 4-1 is  for Tata's trucks (Land Rover).  These vehicles are estimated to have 100
                                            4-35

-------
Regulatory Impact Analysis
g/mi greater CCh 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 CCh lower emissions than the average truck.

       While this comparison confirms the differences in the technology penetrations shown
in Table 4-22, it does not yet explain why these differences exist.  Two well known factors
affecting vehicle fuel efficiency are vehicle weight and performance. The footprint-based
form of the proposed €62 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. Also, higher performing vehicles also tend to have higher CCh emissions over the
two-cycle test procedure. So manufacturers with higher average performance levels will tend
to have higher average CCh emissions for any given footprint. Table 4-23 shows each
manufacturer's average ratios of weight to footprint and horsepower to weight.

                    Table 4-23 Vehicle Weight to Footprint and Performance

Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Car
Weight/
Footprint
(Ib/sq ft)
78
74
73
77
76
67
70
67
73
74
72
82
73
70
78
71
80
73
Horsepower/
Weight
(hp/lb)
0.073
0.054
0.068
0.057
0.057
0.051
0.052
0.05
0.05
0.052
0.059
0.106
0.057
0.049
0.077
0.054
0.059
0.056
Truck
Weight/
Footprint
(Ib/sq ft)
94
85
97
84
83
83
84
79
80
83
80
96
79
81
110
80
108
83
Horsepower/
Weight (hp/lb)
0.059
0.053
0.057
0.052
0.059
0.055
0.056
0.057
0.055
0.056
0.058
0.073
0.054
0.062
0.057
0.062
0.052
0.058
       The impact of these two factors on each manufacturer's "no technology" CCh
emissions was estimated.  First, the "no technology" CCh emissions levels were statistically
                                        4-36

-------
                                        Results of Proposed and Alternative Standards
analyzed to determine the average impact of weight and the ratio of horsepower to weight on
CO2 emissions. Both factors were found to be statistically significant at the 95 percent
confidence level. The results of the statistical analysis are summarized in Table 4-24.

          Table 4-24 Effect of Weight and Performance on "No Technology" Vehicle CO2




Car
Truck
Intercept
(g/mi CO2)


-45.8
-21
Effect of
weight
(g/mi
C02/lb)
0.0819
0.0782
Effect of
Horsepower /
Weight (g/mi
CO2*lb/hp)
1590
1838
R-Square



0.82
0.71
       Together, these two factors explain over 80 percent of the variability in vehicles' CCh
emissions for cars and over 70 percent for trucks. These relationships were then used to
adjust each vehicle's "no technology" CCh emissions to the average weight for its footprint
value and to the average horsepower to weight ratio of either the car or truck fleet, as follows:

       For Cars:

       CO2 Emissions adjusted for weight and performance = "No Technology" CO2 -

              (Vehicle Weight - Vehicle Footprint * 73) * 0.0819 -

              (Vehicle hp/wt - 0.056 ) * 1590

       For Truck:

       CO2 Emissions adjusted for weight and performance = "No Technology" CO2 -

              (Vehicle Weight - Vehicle Footprint * 83) * 0.0782 -

              (Vehicle hp/wt - 0.058 ) * 1838

       We then recomputed the difference between the sales-weighted average of each
manufacturer's adjusted "no technology" CCh levels and their required CCh emission level
under the proposed 2016 standards and subtracted the difference for the industry as a whole.
The results are shown in Figure 4-2.
                                            4-37

-------
Regulatory Impact Analysis
        40
        30
        20
        10
     O -1
     u
       -20
       -30
       -40
               IE
E
H
Figure 4-2 CO2 Emissions Relative to Fleet Adjusted for Technology, Footprint, Weight at Footprint, and
                                     Performance

       First, note that the scale in Figure 4-2 is much smaller by a factor of 3 than that in
Figure 4-1. In other words, accounting for differences in vehicle weight (at constant
footprint) and performance dramatically reduces the differences in various manufacturers'
CO2 emissions.  Most of the manufacturers with high offsets in Figure 4-1 now show low or
negative offsets. For example, BMW's and VW's trucks show very low CCh 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
4-24 are weight and performance. EPA has not determined why some manufacturer's vehicle
weight is relatively high for its footprint value, nor whether this weight provides additional
utility for the consumer.  Performance is more straightforward.  Some consumers desire high
performance and some manufacturers orient their sales towards these consumers.  However,
the cost in terms of €62 emissions is clear.  Producing relatively heavy or high performance
vehicles increases CCh emissions and will require greater levels of technology in order to
meet the proposed CCh standards.
                                        4-38

-------
                                                                 Emissions Impacts

CHAPTER 5: Emissions Impacts

5.1 Overview

    The domestic transportation sector emits approximately 28% of total U.S. greenhouse gas
(GHG) emissions in 2010 based on the standard accounting methodology used by EPA in
compiling the inventory of U.S. GHG emissions pursuant to the United Nations Framework
Convention on Climate Change. This number is potentially even higher, as the standard
methodology excludes upstream transportation fuel emissions associated with extraction,
shipping, refining, and distribution from the emissions of the transportation sector. Within the
transportation sector, emissions from light duty vehicles such as passenger cars, passenger
trucks, and light duty commercial trucks account for 18% of total US GHG emissions, or
approximately 1,300 million metric tons (MMT) of CO2 equivalent (CO2 EQ) greenhouse gas
emissions in calendar year 2010.

    Today's proposal quantifies anticipated impacts from the EPA vehicle CO2 emission
standards. The emissions from the GHGs carbon dioxide (CO2), methane (CH4), nitrous
oxide (N2O) and hydrofluorocarbons (HFCs) were quantified. In addition to reducing the
emissions of greenhouse gases, today's proposal would also influence the emissions of
"criteria" air pollutants, including carbon monoxide (CO), fine paniculate matter (PM2.5) and
sulfur dioxide (SOx) and the ozone precursors hydrocarbons  (VOC) and oxides of nitrogen
(NOx); and air toxics (hazardous air pollutants, including benzene, 1,3-butadiene,
formaldehyde, acetaldehyde, and acrolein).

    Downstream (tailpipe) emission impacts were developed using a spreadsheet analysis
based on data from two EPA models.  Computation algorithms and achieved CO2 levels were
derived from EPA's Optimization Model for reducing Emissions of Greenhouse gases from
Automobiles (OMEGA) and were coupled with non-CO2 emission rates from EPA's Motor
Vehicle Emission Simulator (MOVES).

    Upstream (fuel production and distribution) emission changes resulting from the
decreased fuel consumption predicted by the downstream models were calculated using a
spreadsheet model based on emission factors from GREET.1  Based on these analyses, the
control programs proposed in this chapter would account for  325 MMT CO2EQ of annual
GHG reduction by the year 2030 and 519 MMT per year by 2050. Fuel savings resulting
from the GHG standards are projected at 42.6 billion gallons  of fuel savings in Calendar Year
2050 (Table 5-1).
                                        5-1

-------
 Draft Regulatory Impact Analysis
            Table 5-1 - Impacts of Proposed Program on GHG Emissions and Fuel Savings

CALENDAR
YEAR


2020
2030
2040
2050
ANNUAL GHG
REDUCTION (CO2
EQ MMT)


165.2
324.6
417.5
518.5
FUEL SAVINGS
(MILLION BARRELS
PER DAY OF
GASOLINE
EQUIVALENT)
0.9
1.7
2.2
2.8
ANNUAL FUEL
SAVINGS
(BILLION GALLONS
OF GASOLINE
EQUIVALENT)
13.4
26.2
33.9
42.6
        The emissions of non-GHG air pollutants due to light duty vehicles are also expected
 to be affected by today's proposal. These effects are largely due to changes in fuel
 production, but are also driven by changes in driver behavior.A  The delta values shown here
 include both upstream and downstream contributions.

        Table 5-2 - Impacts of Proposed Program on Non-GHG Emissions (Short Tons per year)
POLLUTANT
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Paniculate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR
YEAR
2020
11.5
16.8
0.2
-83.6
70,614
-28.3
-17,206
-2,856
-16,307
-73,739
%
CHANGE
VS. 2020
REFEREN
CE
0.07%
0.037%
0.00%
-0.04%
0.13%
-0.03%
-0.14%
-0.08%
-0.18%
-0.60%
CALEND
ARYEAR
2030
36.8
60.6
1.8
-77.5
227,832
-15.7
-27,726
-5,431
-31,965
-142,347
%
CHANGE
VS. 2030
REFEREN
CE
0.22%
0.134%
0.03%
-0.04%
0.38%
-0.02%
-0.23%
-0.16%
-0.34%
-1.17%
     We also analyzed the emission reductions over the full model year lifetime of the 2012-
 2016 model year cars and trucks affected by today's proposal.  These results, including both
 upstream and downstream GHG contributions, are presented below (Table 5-3).
A A rebound of 10% is used in this analysis. See section 5.3.3.1.1 for a brief definition of rebound, and the joint
 Technical Support Document for a more complete discussion.
                                           5-2

-------
                                                                   Emissions Impacts
                Table 5-3 - Model Year Lifetime Fuel Savings and GHG Reductions
Model Year
2012
2013
2014
2015
2016
Total
Program
Benefit
Lifetime GHG
Reduction
(MMT CO2 EQ)
81
125
174
243
323
947
Lifetime Fuel Savings
(Billion Gallons Of
Gasoline Equivalent)
6.6
10.0
13.9
19.5
26.3
76.2
Lifetime Fuel Savings
(Million Barrels of
Gasoline Equivalent)
157
239
331
463
626
1,815
5.2 Introduction

5.2.1 Scope of Analysis

    Today's program proposes new standards for the greenhouse gas (GHG) emissions of
light duty vehicles from model year 2012 through model year 2016. The proposed program
affects light duty gasoline and diesel fueled vehicles.  Most passenger vehicles such as cars,
sport utility vehicles, vans, and pickup trucks are light duty vehicles. Such vehicles are used
for both commercial and personal uses and are significant contributors to the total United
States (U.S.) GHG emission inventory. Today's proposal will significantly decrease the
magnitude of these emissions. Because of anticipated changes to driving behavior and fuel
production, a number of co-pollutants would also be affected by today's proposal.

    This chapter describes the development of inventories for emissions of the gaseous
pollutants impacted by the proposed rule.  These pollutants are divided into greenhouse gases,
or gases that in an atmosphere absorb  and emit radiation within the thermal infrared range,
and non-greenhouse gases. Such impacts may  occur "upstream" in the agricultural sector and
fuel production and distribution processes, or "downstream" in direct emissions from the
transportation sector. Table 5-4 presents the processes considered in each domain. This
analysis presents the projected impacts of today's proposal on greenhouse gases in calendar
years 2020, 2030, 2040 and 2050.  Non-greenhouse gases are shown in 2020 and 2030. The
program was quantified as the difference in mass emissions between the proposed standards
and a reference case as described in Section 5.3.2.2.
                                         5-3

-------
Draft Regulatory Impact Analysis
                             Table 5-4 - Processes Considered
PROCESS
Grade Oil Extraction
Grade Oil Transport
Oil Refining
Fuel Transport and Distribution
Fuel Tailpipe Emissions
Air Conditioning System Leakage
UPSTREAM / DOWNSTREAM
Upstream
Upstream
Upstream
Upstream
Downstream
Downstream
    Delta inventories for the four greenhouse gases carbon dioxide (CO2), methane (CH4),
nitrous oxide (N2O) and hydrofluorocarbons (HFC) are presented herein. The sole HFC
discussed in this inventory is R-134a, which is the refrigerant in most current vehicle air
conditioning systems.  Delta inventories for the non-ghg pollutants 1,3-butadiene,
acetaldehyde, acrolein, benzene, carbon monoxide (CO),  formaldehyde, oxides of nitrogen
(NOx), paniculate matter below 2.5 micrometers, oxides of sulfur (SOX), and volatile organic
compounds (VOC) are also presented.

5.2.2 Downstream Contributions

       The largest source of GHG reductions from today's proposal is new standards for
tailpipe emissions produced during vehicle operation (termed "downstream" emissions).
Absolute reductions from tailpipe GHG standards are projected to grow over time as the fleet
turns over to vehicles affected by the standards, meaning the benefit of the program will
continue to grow as long as the  older vehicles in the fleet are replaced by newer, lower CO2
emitting vehicles.

    As described herein, the downstream reductions in greenhouse gases due to the proposed
program are anticipated to be achieved through improvements to both fuel economy and air
conditioning system operation.  Improvements to air conditioning systems can be further
separated into reducing leakage of HFCs (direct improvement) and reducing fuel consumption
by increasing the efficiency  of the air conditioning system (indirect).

    Due to the rebound effect, improving fuel economy is anticipated to increase total vehicle
miles traveled, which has impacts on both GHG and non-GHG emissions. These impacts are
detailed in Section 5.3.3.1.1

5.2.3 Upstream Contributions

       In addition to downstream emission reductions, reductions are expected in the
emissions associated with the processes involved in getting petroleum to the pump, including
the extraction and transportation of crude oil, and the production and distribution of finished
gasoline (termed "upstream" emissions).  Changes are anticipated in upstream emissions due
to the expected reduction in  the volume of fuel consumed. Less gasoline consumed means
less gasoline transported, less gasoline refined, and less crude oil extracted and transported to
                                         5-4

-------
                                                                    Emissions Impacts

refineries. Thus, there should be reductions in the emissions associated with each of these
steps in the gasoline production and distribution process.

    HFC manufacture is not considered a significant source of upstream emissions and is not
considered in this analysis.2

5.2.4 Global Warming Potentials

    Throughout this document, in order to refer to the four inventoried greenhouse gases on
an equivalent basis, Global Warming Potentials (GWPs) are used. In simple terms, GWPs
provide a common basis with which to combine several gases with different heat trapping
abilities into a single inventory (Table 5-5). When expressed in CCh equivalent (CCh EQ)
terms, each gas is weighted by its heat trapping ability relative to that of carbon dioxide. The
GWPs used in this chapter are drawn from publications by the Intergovernmental Panel on
Climate Change (IPCC).3'B
                 Table 5-5 - Global Warming Potentials for the Inventory GHGs
Gas
C02
CH4
N2O
HFC (R134a)
Global Warming potential
(CO2 Equivalent)
1
25
298
1430
5.3 Program Analysis and Modeling Methods

5.3.1 Models Used

       The inventories presented in this document were developed from established EPA
models.

       Downstream inventories were generated using algorithms from EPA's Optimization
Model for reducing Emissions of Greenhouse gases from Automobiles (OMEGA). Broadly
       B The global warming potentials (GWP) used in the NPRM inventory analysis are consistent with
Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4). At this time, the IPCC
Second Assessment Report (SAR) global warming potential values have been agreed upon as the official U.S.
framework for addressing climate change. The IPCC SAR GWP values are used in the official U.S. greenhouse
gas inventory submission to the United Nations climate change framework. When inventories are recalculated
for the final rule, changes in GWP used may lead to adjustments.
                                          5-5

-------
Draft Regulatory Impact Analysis

speaking, OMEGA is used to determine the most likely paths by which manufacturers would
meet tailpipe CCh emission standards. OMEGA applies technologies with varying degrees of
cost and effectiveness to a defined vehicle fleet in order to meet a specified GHG emission
target and calculates the costs and benefits of doing so. The benefits analyses in OMEGA are
conducted in a Microsoft Excel Workbook (the benefits post-processor).  The OMEGA
benefits post-processor produces a national scale analysis of the benefits (emission reductions,
monetized co-benefits) of the analyzed program

       Inputs to the OMEGA post-processor were updated with emission rates from EPA's
Motor Vehicle Emission Simulator (MOVES).45 CO2 emission and fuel consumption rates
are drawn from OMEGA results, with all co-pollutant emission rates derived from the Draft
MOVES emission rate database. c'6  Air conditioning inventories (including HFC and CO2
contributions) were separately calculated in spreadsheet analyses, and are based on previous
EPA research.7 Both MOVES and OMEGA are established models  and continue to be
actively developed.8'9

       Upstream emissions were calculated using the same tools as were used for the
Renewable Fuel Standard 2 (RFS2) proposed rule analysis,10 but for the current analysis it
was assumed that all impacts are related to changes in volume of gasoline produced and
consumed, with no changes in volumes of other petroleum-based fuels, ethanol, or other
renewable fuels. The estimate of emissions associated with production of gasoline from crude
oil is based on emission factors in the GREET model developed by DOE's Argonne National
Lab.11'12  The actual calculation of the emission inventory impacts of the decreased gasoline
production is done in EPA's spreadsheet model for upstream emission impacts. This model
uses the decreased volumes of the crude based fuels and the various crude production and
transport emission factors from GREET to estimate the net emissions impact. As just noted,
the analysis for today's proposal assumes  that all changes in volumes of fuel used affect only
gasoline, with no effects on use of other petroleum-based fuels, ethanol, or other renewable
fuels.

       The following sections provide an in-depth description of the inputs and methodology
used in each analysis.

5.3.2 Description of Scenarios

       One reference and one control scenario are modeled in this proposal, and each is
described below.13 The two scenarios shown are differentiated by their regulatory CO2
emission standards. The reference scenario CO2 emissions are based upon the National
Highway Traffic Safety Administration (NHTSA) Model Year 2011 Corporate Average Fuel
Economy (CAFE) standards,14 while the control scenario CO2 emissions are based upon the
program proposed herein. Otherwise, the scenarios share fleet composition, sales, base
vehicle miles traveled (VMT), and all other relevant aspects. Vehicles are modeled as
compliant with Tier 2 criteria emission standards.
 Two tables were updated in the Draft MOVES database. These tables are available in the docket.
                                         5-6

-------
                                                                    Emissions Impacts

       Ethanol use was modeled at the volumes projected in AEO2007 for the reference and
control case; thus no changes are projected in upstream emissions related to ethanol
production and distribution. However, due to the decreased gasoline volume associated with
today's proposal, a greater market share of E10 is expected relative to EO, which would be
expected to have some effect on fleetwide average non-GHG emission rates. The increased
market share of ethanol blended gasoline, which is likely small relative to the other effects
considered here,  has not been accounted for in the downstream emission modeling conducted
for today's proposal, but is planned to be  addressed in the final rule air  quality analysis, for
which localized impacts could be more significant.  Due to the lower energy content of
ethanol blended gasoline, the increase in  ethanol market share is also projected to decrease the
fuel savings predicted by this analysis by approximately 1-2%.  A more comprehensive
analysis of the impacts of different ethanol and gasoline volume scenarios is being prepared
as part of EPA's RFS2 rulemaking package.0

    5.3.2.1  Sales and Fleet Composition

       Fleet composition has a significant effect upon the impacts of the proposed rule.
Consequently, it is significant that the cars and trucks in this analysis are defined differently
than their historic EPA classifications. Passenger Automobiles (PA), as used herein, are
defined as classic cars and two-wheel drive SUVs below 6,000 Ibs. gross vehicle weight.  The
remaining light duty fleet is defined as Non-Passenger Automobiles (NPA). The NPA
classification includes most classic light duty trucks such as four-wheel drive SUVs, pickup
trucks, and similar vehicles.

       As shown in Table 5-6, the vehicle classifications used herein are consistent with the
definitions used by the National Highway Safety Transit Association in the MY 2011 CAFE
standards.15  While the formal definitions are lengthy, brief summaries of the classifications
are shown here.

                           Table 5-6 -Definitions of Vehicle Classes
REGULATOR
National Highway Traffic
Safety Administration CAFE
Program (pre-MY 2011)
EPA Program
(MY 20 12+)
CAR DEFINITION
Classic Car — Passenger Car
Passenger Automobile — PC
+ 2 wheel drive SUVs below
6,000 GVW
TRUCK DEFINITION
Classic Truck — Light Duty Trucks 1 -4 and
Medium Duty Passenger Vehicles.
Non-Passenger Automobile — Remaining
light duty fleet
       Based on EPA analysis of the projected MY 2012-2016 fleet,16 approximately 22% of
the classic truck fleet is anticipated to be reclassified as Passenger Automobiles under the new
standards. Projected sales of classic cars and trucks for calendar years 2012-2030 were drawn
from the Energy Information Administration Annual Energy Outlook (AEO) April 2009
projection.17 The AEO 2009 sales projections, based on the classic fleet, were then
D XX [Insert RFS2 NPRM reference, since FRM will not be available in time for GHG NPRM.]
                                         5-7

-------
Draft Regulatory Impact Analysis
reclassified using PA and NPA definitions.  For calendar years 2030-2050, which are beyond
the scope of AEO's projections, 0.76% annual growth in the sales of cars and trucks was
assumed.

                  Table 5-7 - Projected Total Vehicle Sales and Car Fractions

Total Light
Duty Sales
Classic Car
Fraction
PA Fraction
PAs Sold
Model Year
2012
14,850,955
51.8%
62.6%
8,235,204
Model Year
2013
15,653,713
52.9%
63.6%
9,255,624
Model Year
2014
16,216,393
54.3%
64.3%
9,977,341
Model Year
2015
16,575,580
55.8%
65.6%
10,479,350
Model Year
2016
16,581,055
57.1%
66.5%
10,890,967
    5.3.2.2  Proposed Standards

       Individual PA and NPA tailpipe CO2 fleet average emission standards are shown for
reference and control scenarios along with an anticipated fleet average combined standard.
Rather than an absolute standard, the values referred to here as standards are the production
weighted average standard predicted by the coefficients of the relevant equation. As
documented in Preamble section II, under both reference and  control scenarios, each
manufacturer has a unique fleet average standard based on their vehicle footprints and
production.

       Fleet average standards are calculated here by weighting the individual PA and NPA
standards by the respective proportions of anticipated production (Section 5.3.2.1). These
CO2 emission values are unadjusted values (i.e. in CAFE space), so they are lower than the
anticipated on-road emissions.  In all scenarios, vehicles are assumed to maintain model year
2016 emissions for post-2016 vehicles. Because the fleet composition continues to change
post-MY 2016, the fleet average emission level continues to vary

5.3.2.2.1 Reference Case

        5.3.2.2.1.1  CO2 Emission Standards

       The reference scenario standards were derived from the NHTSA model year 2011
Corporate Average Fuel Economy (CAFE) standards applied  to the MY 2012-2016 reference
fleet (see chapter 1 of the draft joint TSD  and chapter 4 of this DRIA).18 Average car and
truck fuel economy standards were calculated from the coefficients in the MY 2011  rule and
EPA analysis of the projected MY 2012-2016 fleet.19'20 Average fuel economy levels were
calculated for each manufacturer's fleet, and then combined based on projected sales.

-------
                                                                   Emissions Impacts
       A ratio of 8887 grams of CCh emitted per gallon of gasoline was used to convert to the
calculated fuel economy standards to CCh (gram/mile) emission factors. The basic derivation
of the 8887 factor can be seen in previous EPA publications.21

       CO2 emission standards were calculated by applying the CAFE coefficients to the
footprint of each model, and calculating fleet averages based on projected model sales.
Minor changes in the emission standard are expected due to projected changes in the average
new vehicle footprint between 2012 and 2016 (Table 5-8).

             Table 5-8 — Reference Case Average Emission Standards (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
2015
2016
PA
EMISSION
LEVEL
291
291
291
291
291
NPA EMISSION
LEVEL
366
366
368
368
368
MY EMISSION
LEVEL
319
319
319
318
317
        5.3.2.2.1.2  Achieved CC>2 Emission Levels

       The emission standards shown in Table 5-8 do not reflect the impact of several
program flexibilities in CAFE, nor do they account for manufacturer overcompliance.
Projected achieved emission levels include the effects of manufacturers who pay fines rather
than comply with the emission standards, as well as a number of credit programs under
EPCA/EISA that allow manufacturers to emit more than the standard seemingly requires.
Additionally, some manufacturers overcomply with the standards, and this overcompliance is
not reflected in the CAFE standards.

       While the CAFE program is complex, the most significant portions of the program
flexibilities are accounted for. In this analysis, manufacturer overcompliance, credit trading,
FFV credits, and fine paying manufacturers were accounted for. Banked credits from the
calculation of achieved standards were excluded.

       In general, achieved emission levels were estimated by beginning with the more
stringent of either (A) manufacturer's CAFE score (in CCh space) or (B) estimated achieved
MY 2008 CO2 level based on the EPA fleet data file. Using that starting point, each
manufacturer's emissions was increased by the impact of the credits of which is anticipated
that they will take advantage.  Consistent with the use of the MY 2011 standards, the credits
and trading levels available for MY 2011 are assumed available in all years of the reference
case.  Manufacturers were always assumed to perform at least as well as they did in 2008.

Overcompliance and Credit Trading
                                         5-9

-------
Draft Regulatory Impact Analysis
       Using the EPA fleet file, the fleet mix was estimated by manufacturer for model year
2012 through model year 2016. For each model year, the CAFE score (in CCh space) was
calculated by manufacturer for PA and NPA separately. To estimate the effects of
overcompliance, each manufacturer's achieved 2008 PA/NPA emissions were compared
against the PA/NPA emissions required by CAFE in 2011.

       The overcompliance on either PA or NPA could be "traded" within a manufacturer in
order to make up a shortfall in the remaining vehicle class.  Credits are generated on a sales
and VMT weighted basis, and traded between vehicle classes. The MY 2011  CAFE cap on
credit trading of 1.0 mpg was used. This trading of the overcompliance credit negates some,
but not all of the overcompliance anticipated. Certain manufacturers, such as Toyota and
Honda, overcomply by a great deal more than they are able to trade between vehicle classes.

Flex Fueled vehicle Credits

       The 2007 Energy Independence and Security Act allows for CAFE credits due to
production of "flex-fueled" vehicles.  Under the model year 2011 standards, such credits can
be used to meet up to  1.2 MPG of the CAFE standard. The manufacturers General Motors,
Chrysler and Ford were assumed to take advantage of this credit for both cars and trucks,
while Nissan was assumed to utilize this credit solely for trucks.

Fines

       In this analysis, EPA used estimates of fine paying manufacturers from NHTSA's
Volpe model. That model supplied projected maximum stringencies that a manufacturer
would meet before it was more cost effective  to pay a non-compliance fine. The
manufacturers who are projected to pay fines  are Tata, Daimler, BMW, Porsche, and
Volkswagen.

       The projected impacts of these program flexibilities on the standards, achieved levels
based on program flexibilities and manufacturer overcompliance are shown in Table 5-10.

                      Table 5-9 - Impacts of credits (grams/mile CO2 EQ)
MODEL
YEAR
2012
2013
2014
2015
2016
OVERCOMPLIANCE,
CREDITS AND
TRADING
-6.1
-6.4
-6.8
-7.1
-7.3
FFV
6.7
6.7
6.7
6.5
6.4
FINES
0.7
0.2
0.1
0.1
0.0
NET
1.2
0.5
0.0
-0.5
-0.9
                                        5-10

-------
                                                                    Emissions Impacts
                Table 5-10 — Reference Case Achieved Emissions (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
284
283
283
282
282
ANTICIPATED
NPA EMISSION
LEVEL
382
382
383
384
384
ANTICIPATED
MY EMISSION
LEVEL
320
319
319
317
316
5.3.2.2.2  Control Case

         5.3.2.2.2.1  CO2 Emission Standards

       Similar to the reformed CAFE program, EPA is proposing to establish a footprint
attribute based function in order to determine the CCh (gram/mile) emission standard for a
given vehicle.  The piecewise linear function used by EPA is documented in Section II of the
preamble. Based on this function, and the same vehicle fleet as was used in the reference
scenario, EPA calculated projected PA and NPA fleet average emission standards for the
MY2012-2016 vehicles. Average PA and NPA fuel economy standards were calculated by
applying this function to the EPA fleet file (Table 5-11).22

             Table 5-11 - Control Case Average Emission Standards (grams/mile CO2)
MODEL
YEAR


2012
2013
2014
2015
2016
PA
EMISSION
LEVEL

261
253
246
235
223
NPA EMISSION
LEVEL


351
341
332
317
302
PROJECTED
MY EMISSION
STANDARD
LEVEL
295
286
276
263
250
                                         5-11

-------
Draft Regulatory Impact Analysis
        5.3.2.2.2.2  Achieved CO2 Emission Levels

       Just as with the reference scenario, the emission standards (Table 5-11) do not include
the effect of several program flexibilities built into the EPA program.

       The same basic methodology was used to calculate achieved fleet standards for the
control case.  In general, achieved standards were estimated by beginning with the more
stringent of either (A) manufacturer's calculated footprint-based emission standard or (B)
estimated achieved CCh level based on the EPA fleet data file. Using that starting point, each
manufacturer's emissions were increased by the impact of the credits of which were
anticipated that they will take advantage. Manufacturers were always assumed to perform at
least as well as they did in 2008.

Overcompliance and Trading

       Using the EPA fleet file, the fleet mix was estimated by manufacturer for model year
2012 through model year 2016. For each model year, the GHG standard was calculated by
manufacturer for PA and NPA separately. To estimate the effects of overcompliance, each
manufacturer's achieved PA/NPA emissions was compared against the PA/NPA emissions
required by CAFE.

       The achieved overcompliance on either PA or NPA could be "traded" within a
manufacturer in order to make up a shortfall in the remaining vehicle class.  Credits are
generated on a sales and VMT weighted basis, and traded between vehicle classes. Under the
EPA proposed program, there are no limits on such trading. This trading of the
overcompliance credit negates nearly all of the overcompliance anticipated in the early years.

       Under the unlimited within-fleet trading allowed under the EPA program,
manufacturers can potentially invest in their fleet differently than the precise achieved levels
described here.  Because the credit trading is VMT weighted, the resulting changes will be
essentially environmentally neutral on both  GHG and criteria pollutants.

Flex Fueled Vehicles

       The flex fueled vehicle credit, per the discussion in the preamble (Section III), is set at
1.2 MPG for MY 2012-2014, 1.0 MPGforMY 2015, andO MPGforMY 2016+. As in the
reference case, it was assumed that the manufacturers General Motors, Chrysler and Ford
would take advantage of this credit for both cars and trucks, while Nissan would utilize this
credit solely for trucks.

A/C

       Indirect A/C credits were set at 5.7 grams CCh per mile for the fleet, while direct A/C
credits were set at 6.9 grams CCh per mile for PA and 8.6 grams CO2 per mile for NPA).
EPA assumed market penetration of the technology according to Table 5-16. A more
                                        5-12

-------
                                                                   Emissions Impacts

complete discussion of the A/C credit program and inventories is provided in section 5.3.3.2,
as well as DRIA chapter 2.

Temporary Lead Time Allowance Alternative Standards (TLAAS)

       We assumed that every potentially eligible manufacturer took advantage of the
temporary lead time allowance.  Each qualifying manufacturer was assumed to use the full
vehicle allocation according to the default production schedule shown in Preamble Section III.
The allocation was split evenly between cars and trucks for each manufacturer.  This vehicle
allocation was assumed to emit as much CCh per mile as the highest emitting car or truck in
each manufacturer's fleet. These vehicles were then proportionally averaged into the
manufacturer's GHG score. For more on the TLAAS program, please see Appendix A to this
RIA chapter.

       The aggregate impacts of these program flexibilities are listed in Table 5-12.

       Table 5-12 — Estimated Impacts of Proposed Program Flexibilities (grams/mile CO2 EQ)
MODEL
YEAR
2012
2013
2014
2015
2016
OVERCOMPLIANCE,
CREDITS AND
TRADING
0.1
-0.3
0.0
0.0
0.0
FFV
6.0
5.7
5.4
4.1
0.0
DIRECT
A/C
1.9
3.0
4.1
5.6
6.3
INDIRECT
A/C
1.4
2.3
3.1
4.3
4.8
TLAAS
0.3
0.2
0.2
0.1
0.0
NET
9.6
10.9
12.7
14.0
11.1
       Based on these impacts, the achieved emission level by PA, NPA and fleet are
displayed in Table 5-13. Please note that the achieved emission levels include the increase in
test procedure emissions due to the use of the A/C credit.  The impacts of A/C improvements
are discussed in section 5.3.3.2.

         Table 5-13 — Federal GHG Program Anticipated Emission Levels (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
266
261
256
246
234
ANTICIPATED
NPA EMISSION
LEVEL
368
359
349
335
314
ANTICIPATED
MY EMISSION
LEVEL
304
297
289
277
261
       Table 5-13 differs slightly from the OMEGA cost-side model results in 2016.
OMEGA assumes environmentally neutral trading between PA and NPA within a
manufacturer's fleet in order to minimize technology costs. Consequently, the distribution of
fleet emission reductions differs slightly between cars and trucks from that which is shown
                                         5-13

-------
Draft Regulatory Impact Analysis

here.  However, because the trading is VMT weighted, it is environmentally neutral and has
no GHG emissions impacts.

      OMEGA also predicts slight undercompliance in 2016 for several manufacturers, while
the results presented here assume full compliance. Based on preliminary analysis, the
OMEGA cost-side results are estimated to produce approximately 0.5% less GHG benefit.

5.3.3 Calculation of Downstream Emissions

       The fleet inputs (achieved CO2 emission levels by model year and vehicle
sales) described above were incorporated into a spreadsheet along with emission rates
derived from Draft MOVES 2009 and benefits calculations from the OMEGA post-
processor. The resulting spreadsheet projects emission impacts in each calendar year.
The effects of the program grow over time as the fleet turns over to vehicles subject to
the more stringent new standards.

       A model year lifetime analysis, considering only the five model years
regulated underneath the program, is shown in Section 5.6.  In contrast to the calendar
year analysis, the model year lifetime analysis shows the lifetime impacts of the
program on each MY fleet over the course of that fleet's existence.

    5.3.3.1  Tailpipe GHG Emissions

        Two basic elements feed into OMEGA's calculation of vehicle tailpipe
emissions. These elements are VMT and emission rates.

                       Total Emissions = VMT n^ * Emission rate grams/mile

                                    Equation 3 - Emissions

       This equation is adjusted in calculations for various emissions, but provides the basic
form used throughout this analysis. As an example, in an analysis of a single calendar year,
the emission equation is repeatedly applied to determine the contribution of each model year
in the calendar year's particular fleet.  Appropriate VMT and emission factors are applied to
each model year within the calendar year.  Emissions are then summed across all model years.

       The following sections describe the VMT and emission factor components of this
analysis.

5.3.3.1.1 Base VMT

       The downstream analysis is based upon a "bottom-up" estimate of total VMT
and vehicle population.  The VMT inputs are documented more fully  in draft joint
TSD chapter 4, but a description of their use in the emissions calculations are provided
below.

       The analysis spreadsheet contains MY-specific estimates of per vehicle VMT
by vehicle age, as well as the fractions of new vehicles still on the road as a function


                                        5-14

-------
                                                                   Emissions Impacts
of age. The total VMT for vehicles in a specific model year during a specific calendar
year is determined by multiplying 1) new vehicle sales for that model year, 2) the
fraction of new vehicles remaining on the road according to the age of those vehicles
in that calendar year and 3) the annual VMT for that model year, age, and vehicle
class.

       Future vehicle sales were drawn from AEO 2009 (as discussed in Section
5.3.2.1), while historic vehicle sales are drawn from the Transportation Energy Data
Book,23 Post MY 2011 vehicles were reclassified in order to correspond to the
PA/NPA definitions.

       As described in the draft technical support document, mileage accumulation by
age was calculated using inputs from the NHTSA "Vehicle Survivability and Travel
Mileage Schedules" and additional inputs unique to this analysis.24'25 In brief, a
1.15% per vehicle annual VMT growth rate was assumed, but additional factors such
as achieved fuel efficiency and the price of gasoline also contributed to the precise
schedule for each MY.

       The survival schedule was taken without emendation from "Vehicle
Survivability  and Travel Mileage Schedules." While adjustments may be necessary to
this schedule  to accommodate the change between classic cars/trucks and PA/NPA,
EPA is unaware of any extant data supporting specific adjustments.  Because of the
lack of data, the survival rates from "Vehicle Survivability and Travel Mileage
Schedules" were used without further adjustment (Table 5-14).26
                                         5-15

-------
Draft Regulatory Impact Analysis
                           Table 5-14 - Survival Fraction by Age
AGE
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
PA SURVIVAL
FRACTION
0.9950
0.9900
0.9831
0.9731
0.9593
0.9413
0.9188
0.8918
0.8604
0.8252
0.7866
0.7170
0.6125
0.5094
0.4142
0.3308
0.2604
0.2028
0.1565
0.1200
0.0916
0.0696
0.0527
0.0399
0.0301
0.0227
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
NPA
SURVIVAL
FRACTION
0.9950
0.9741
0.9603
0.9420
0.9190
0.8913
0.8590
0.8226
0.7827
0.7401
0.6956
0.6501
0.6042
0.5517
0.5009
0.4522
0.4062
0.3633
0.3236
0.2873
0.2542
0.2244
0.1975
0.1735
0.1522
0.1332
0.1165
0.1017
0.0887
0.0773
0.0673
0.0586
0.0509
0.0443
0.0385
0.0334
                                        5-16

-------
                                                                   Emissions Impacts
       A complete discussion of the derivation of the MY specific VMT schedules is
provided in draft joint TSD chapter 4.

5.3.3.7.2 Rebound

       The tailpipe €62 standards are expected to result in greater fuel efficiency. Per
the discussion of the rebound effect in the draft joint TSD chapter 4, improved fuel
efficiency is expected to lead to a proportional increase in VMT. Consequently, the
VMT differs between the reference and control cases.

       The rebound effect is formally defined as the ratio of the percentage change in
VMT to the percentage change in incremental driving cost, which is typically assumed
to be the incremental cost of fuel consumed per mile. Since VMT increases with a
reduction in fuel consumption, the sign of the rebound effect is negative. The
percentage increase in VMT for a given change in fuel consumption per mile is
calculated as follows:
                             = -BEB *
                                                Flee,FCM
                                                              Equation 4 - VMT Rebound
       As fuel consumption changes by model year, each model year's vehicles
reflect a different change in VMT  In OMEGA, this change in VMT is assumed to
continue throughout the life of the vehicle, which is consistent with the assumption
that fuel economy is constant throughout vehicle life.

       This analysis assumes a 10% rebound effect; the analysis behind 10% is
explored in greater depth in Chapter 4 of the draft joint TSD.

5.3.3.1.3 Emission Factors

       The derivation of the emission factors used in this analysis is documented in chapter 4
of the technical support document. Briefly, CCh emission rates are derived from the achieved
vehicle emission levels in Table 5-10 & Table 5-13,  SCh emission rates are derived from fuel
sulfur levels, and the emission rates for the remaining pollutants are derived from the draft
MOVES 2009 database.  For a more complete discussion of these emission rates, please see
the draft joint TSD chapter 4.27

       EPA is not projecting any reductions in tailpipe CH4 or N2O emissions as a result of
these proposed emission caps, which are meant to prevent emission backsliding and to bring
diesel vehicles equipped with advanced technology aftertreatment into alignment with current
gasoline vehicle emissions. Similar to other pollutants, there are emission impacts due to
reduced fuel production and increased driving (rebound).
                                        5-17

-------
Draft Regulatory Impact Analysis
5.3.3.1.4  Tailpipe CO2 Emissions from Vehicles

       CCh emission rates were derived from the achieved levels in Table 5-10 &
Table 5-13. Previous EPA analysis has shown that an approximately 20% gap exists
between CAFE space fuel economy and on-road fuel economy.28 The on-road gap is
more fully documented in the draft joint TSD chapter 4.

       The 20% gap, while approximate, includes average effects of fuel efficiency
contributors such as road roughness, wind, and high acceleration events. The gap also
reflects the different energy content between certification fuel and real world fuel
(which frequently contains some oxygenate or ethanol.), as well as the fuel
consumption impacts of running a mobile vehicle air conditioning system.  In this
analysis, CCh emissions are assumed to remain constant throughout the  vehicle's
lifetime.

       By dividing a CAFE-space CCh emission rate by  (1-on-road gap), one can
approximate the actual on-road CCh emissions experienced by drivers.

       On road tailpipe CO2 emissions =
       Achieved CO2 Emission Level /(1-on-road gap) x VMT including rebound
                      Equation 5- Tailpipe CO2 Emissions Excluding A/C

       Based on Equation 5, the baseline CCh emissions and change in  tailpipe
emissions due to the new control program were calculated. Emissions due to rebound
were also calculated. The contributions of the A/C control program are  excluded from
this table.

             Table 5-15 - Tailpipe CO2 Emissions including Baseline A/C Usage (MMT)

Tailpipe CCh Emissions
(Reference)
A CCh Emissions
(Control) including 10%
rebound
A CO2 Emissions due to
10% rebound
2020
1,209
-108
11
2030
1,373
-212
21
2040
1,662
-274
27
2050
2,069
-344
33
    5.3.3.2  Air Conditioning Emissions

       Outside of the tailpipe CCh emissions directly attributable to driving, EPA has
analyzed how new control measures might be developed for air conditioning (" A/C")-related
emissions of HFCs and CCh. With regard to air conditioning-related emissions, significant
opportunity exists to reduce HFC emissions from refrigerant leakage (direct emissions) and
CCh from A/C induced engine loads (indirect emissions).
                                        5-18

-------
                                                                   Emissions Impacts

       Over 95% of the new cars and light trucks in the U.S. are equipped with A/C systems.
There are two mechanisms by which A/C systems contribute to the emissions of GHGs. The
first is through direct leakage of refrigerant (currently the HFC compound R134a) into the air.
Based on the high GWP of HFCs (Table 5-5), a small leakage of the refrigerant has a greater
global warming impact than a similar amount of emissions from other mobile source GHGs.
Leakage can occur slowly through seals, gaskets, hose permeation and even small failures in
the containment of the refrigerant, or more quickly through rapid component deterioration,
vehicle accidents or during maintenance and end of-life vehicle scrappage (especially when
refrigerant capture and recycling programs are less efficient). The leakage emissions can be
reduced through the choice of leak-tight, durable components, or the global warming impact
of leakage emissions can be  addressed by using an alternative refrigerant with lower GWP.
These options are described  more fully in DRIA Chapter 2.

       EPA's analysis indicates that together, these A/C- related emissions account for
approximately 8% of the GHG emissions from cars and light trucks. EPA is proposing credit
provisions which we expect  all manufacturers to utilize which are expected to reduce direct
leakage emissions by 50% and to reduce the incremental increase of A/C related CCh
emissions by 40% in model year 2016 vehicles, with a gradual phase-in starting in model year
2012. It is appropriate to separate the discussion of these two categories of A/C-related
emissions because of the fundamental differences in the emission mechanisms and the
methods of emission control. Refrigerant leakage control is akin in many respects to past EPA
fuel evaporation control programs in that containment of a fluid is the key control feature,
while efficiency improvements are more similar to the vehicle-based control of CCh in that
they would be achieved through specific hardware and controls.

       The anticipated phase-in of air conditioning controls is shown in Table 5-16. The
market penetration is based upon analysis from the OMEGA model. OMEGA projections
show improved A/C technology market penetration at 85% of the market in 2016. This 85%
cap is then roughly linearized across the five year period (Table  5-16).  Because HFC leakage
is somewhat independent of vehicle miles traveled, HFC reduction % is based on the
proportion of new vehicles that have HFC leakage containment technology. By contrast,
indirect A/C reduction % is dependent upon the travel fraction, and is proportional to the
VMT attributable to vehicles with the control technology.

             Table 5-16 - AC Control by Model Year (Reduction from Base Emissions)

Market Penetration of technology
HFC Reduction %
Indirect Reduction %
MY
2012
25%
-13%
-10%
MY
2013
40%
-21%
-16%
MY
2014
55%
-28%
-22%
MY
2015
75%
-38%
-30%
MY
2016+
85%
-43%
-34%
       The penetrations of A/C control technology and HFC reductions shown in this section
differ slightly from those shown in DRIA chapter 2, and in Preamble section III.  The HFC
credits discussed in this section are slightly larger (9%) than the proposed program credits,
while the penetration schedule of the credits is slightly lower. The net effect is a less than 1%
                                        5-19

-------
Draft Regulatory Impact Analysis
overstatement of total program GHG related benefits. EPA will address this issue in the final
rulemaking.
5.3.3.2.7 Direct A/C (HFC) Emissions
       The projected HFC baseline inventories are derived from previous EPA analyses.
29
       HFC emissions are a leakage type emission, similar to other evaporative emissions
from a vehicle.30 Consequently, HFC emissions are tied more closely to vehicle stock than to
VMT.

       To calculate HFC emissions, the per-vehicle per-year emission contribution of the
current vehicle fleet was determined using averaged 2005 and 2006 registration data from the
Transportation Energy Databook (TEDB)31 and 2005 and 2006 mobile HFC leakage estimates
from the EPA Emissions and Sinks report.  This per-vehicle per-year contribution was then
scaled to the projected vehicle fleet in each future year using data from the emission modeling
analysis. This analysis assumes that the leakage rates of the current fleet remain constant into
the future. Preliminary EPA estimates suggest that air conditioner charge size is decreasing,
which implies that the current analysis may somewhat overstate the HFC emission inventory.

       The resulting HFC  inventory is a combination top-down/bottom up inventory and
includes leakage, maintenance/servicing, and disposal/end of life phases of HFC. The
proposed EPA program is expected to impact only two of these phases of the HFC inventory
by reducing leakage and reducing need for servicing.

       The vehicle population model from  the emission analysis was used to calculate the
penetration of the technology into the market by calendar year.  The equation used for
calculating the reductions in HFC is shown below (Equation 6).

Emissions Reductions = Reduction % by Calendar Year x Total CY inventory
Reduction % by CY = ^calendar Year (Reduction % by MY x Vehicle Population by MY)/Total Vehicle Population
                          Equation 6 - HFC Inventory Calculation

       Table 5-17 shows the calculated penetration of technology into the vehicle fleet and
consequent reduction from baseline HFC Inventory.

                          Table 5-17  - HFC (Direct A/C) Emissions
Calendar
Year
2010
2020
2030
2040
2050
Baseline HFC
(MMT CO2EQ)
55.0
61.5
70.5
76.4
82.1
Penetration of
Technology
(Population
Based)
0%
43%
76%
84%
85%
Reduction From
Baseline
(%)
0%
-21%
-38%
-42%
-42%
Reduction from
Baseline
(MMT CO2EQ)
0.0
-13.5
-27.2
-32.1
-34.9
                                         5-20

-------
                                                                   Emissions Impacts
5.3.3.2.2 Indirect A/C (CO2) Emissions

       By adding an additional load to the powertrain, A/C indirectly causes an increase in
tailpipe CCh emissions. Thus, where HFC inventory is proportional to penetration of the
technology into the vehicle population, the indirect A/C emission inventory is proportional to
VMT of those vehicles. Because newer vehicles are assumed to be driven more, indirect A/C
control technology benefits the fleet more quickly than HFC control technology.

       The emission rates for indirect A/C usage were taken from the EPA analysis
documented in DRIA chapter 2.  There, indirect A/C usage is calculated to add 14.25 grams
of CO2 emissions to the certification emissions of either cars or trucks.  The indirect  A/C
controls proposed in the rule are estimated to remove up to 40% of the emission impact of air
conditioning systems, or 5.7 grams per mile.

       The OMEGA post processor was used to calculate the contribution the indirect A/C
program to the overall inventory. Reference and Control scenario emissions due to anticipated
improvements to indirect A/C systems are shown in Table 5-18.

                            Table 5-18 —Indirect A/C Emissions
Calendar
Year
2010
2020
2030
2040
2050
Baseline Indirect A/C
(MMT CO2EQ)
52.9
55.2
65.9
80.6
100.5
Reduction From
Baseline
(%)
-0%
-20%
-32%
-34%
-34%
Reduction from
Baseline
(MMT CO2EQ)
0
-11.0
-21.1
-27.2
-34.1
       It should be noted that the baseline indirect A/C emissions are included within the on-
road adjustment factor. The baseline inventory is not double counted when aggregating the
components of this program.

    5.3.3.3  Tailpipe Co-pollutant Emissions

       Due to the rebound effect, the downstream emissions of several co-pollutants are
anticipated to increase. These inventories are calculated in a similar manner to the CCh
emissions. Rebound VMT, which is the additional driving, is broken into distribution by
vehicle age.  VMT by each age was then multiplied by the appropriate emission factor. These
emissions by age were then summed by calendar year (Equation 7,Table 5-19).

       The EPA reference fleet assumes a small  number of diesel vehicles are sold in each
year (~20 thousand out of-13-16 million).  For the criteria emission analysis, it was assumed
that 0.5% of new light duty vehicles sold were diesels. Because diesel fueled vehicles are
                                         5-21

-------
Draft Regulatory Impact Analysis

subject to the same Tier 2 emission standards as gasoline fueled vehicles, the emission rates
of criteria pollutants are similar.E
           EmissionscaiendarYear = Zcaiendar Year (Rebound VMT by Age * Emission Factor by Age )
                              Equation 7 - Emissions by Calendar Year
                  Table 5-19 - Delta GHG Emissions Due to Rebound (Metric Tons)


Gasoline Fueled Vehicles
ACH4
AN2O

Diesel Fueled Vehicles
ACH4
AN2O
CALENDAR YEAR
2020

307
134


0.13
0.20
2030

644
284


0.25
0.40
2040

857
377


0.32
0.52
2050

1079
3475


0.40
0.66
E Emissions rates between tier 2 gasoline and diesel vehicles are similar but not identical due to the particulars of
operations of the engine types. Diesel and gasoline engines emit differently during start, as well as during the
various modes of operation.
                                              5-22

-------
                                                                   Emissions Impacts
                Table 5-20 - Delta Downstream non-GHG Emissions (Short Tons)


Gasoline Fueled Vehicles
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A CO
A Formaldehyde
ANOX
A PM 2.5
ASO2
AVOC

Diesel Fueled Vehicles
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A CO
A Formaldehyde
ANOX
A PM 2.5
AS02
AVOC
CALENDAR YEAR
2020

13.08
24.58
1.22
79.27
77,648.80
31.08
5,128.09
217.31
-2,502.6
1,678.90


0.17
0.23
0.07
0.37
174.13
0.72
225.72
0.88
2030

39.75
74.86
3.72
241.26
241,314.67
94.43
15,110.31
569.82
-4,905.6
5,442.80


0.44
0.60
0.17
0.98
624.39
1.89
449.95
1.87
Attributed to Gasoline
19.33
51.03
       In general, downstream emissions are predicted to increase a small amount due to
rebound driving. The one exception is sulfur emissions (SO2), which are predicted to
decrease  as a result of the decrease in fuel consumption.  As shown in section 5.3.4, the
increases in non-ghg pollutants are generally less than the projected decreases on the upstream
side. The exceptions are in those pollutants, such as carbon monoxide (CO), where a
relatively small of US emissions comes from upstream sources.

    5.3.3.4  Fuel Savings

       The proposed EPA  program is anticipated to create significant fuel savings as
compared to the reference case.  Projected fuel savings are shown in Table 5-21. Fuel savings
can be calculated from total tailpipe CO2 avoided (including CO2 due to driving and indirect
                                        5-23

-------
Draft Regulatory Impact Analysis
A/C use) using a conversion factor of 8887 grams of CCh per gallon of gasoline. All fuel
saved is considered 100% gasoline without any oxygenate/'
                            F,32
       Fuel savings were calculated from total tailpipe CO2 avoided (including CO2 due to
driving and indirect A/C) using a conversion factor of 8887 grams of CCh per gallon of
gasoline. 33

              Table 5-21 - Downstream Fuel Consumption Changes by Calendar Year
                          (Billions of Gallons of Gasoline Equivalent)

Fuel Consumption (Reference)
A Total Fuel Consumption due
to EPA Program
A Fuel Consumption due to
10% rebound
A Fuel Consumption due to
A/C controls
2020
142.2
-13.4
1.2
-1.2
2030
161.9
-26.2
2.3
-2.4
2040
196.2
-33.9
3.0
-3.1
2050
244.1
-42.6
3.8
-3.8
5.3.4 Calculation of Upstream Emissions

       The term "upstream emissions" refers to air pollutant emissions generated from all
crude oil extraction, transport, refining, and finished fuel transport, storage, and distribution.
As shown above in Table 5-4 this includes all the stages prior to the final filling of vehicle
fuel tanks at retail service stations.  The details of the assumptions, data sources, and
calculations that were used to estimate the emission impacts presented here can be found in
the Technical Support Document and the docket memo, "Calculation of Upstream Emissions
for the GHG Vehicle Rule.""
34
5.4 Greenhouse Gas Emission Inventory

       This section presents total program calendar year impacts by sector (Table 5-22, Table
5-23,Table 5-24). Upstream, downstream, and total program impact are presented.
F Based on the documentation of the on-road gap, it would be justifiable to assume an ethanol percentage of
0.8%.  This volume of ethanol would result in a total energy difference of less than 0.5%. See the fuel labeling
rule technical support document, EPA420-R-06-017, for further details.
                                         5-24

-------
                                                                 Emissions Impacts
  Table 5-22 - Downstream GHG and Fuel Consumption Changes vs. Reference Case

A CO2 (Metric Tons)
A CH4 (Metric tons)
A N2O (Metric tons)
A HFC (Metric tons)
A GHG (MMT CO2 EQ)
A Fuel Consumption (billion
gallons per year)
2020
-118,682,739
308
134
-9,429
-132.1
-13.4
2030
-232,643,716
645
284
-18,987
-259.7
-26.2
2040
-301,498,777
857
378
-22,420
-333.4
-33.9
2050
-378,287,357
1,080
476
-24,407
-413.0
-42.6
               Table 5-23 - Upstream GHG Change vs. Reference Case

A CO2 (Metric Tons)
A CH4 (Metric tons)
A N2O (Metric tons)
A GHG (MMT CO2 EQ)
2020
-28,857,236
-163,638
-464
-33.1
2030
-56,566,395
-320,765
-909
-64.9
2040
-73,308,230
-415,701
-1,178
-84.1
2050
-91,979,068
-521,576
-1,478
-105.5
      Table 5-24 - Total GHG and Fuel Consumption Changes vs. Reference Case

A CO2 (Metric Tons)
A CH4 (Metric tons)
A N2O (Metric tons)
A HFC (Metric tons)
A GHG (MMT CO2 EQ)
A Fuel Consumption
(billion gallons per year)
2020
-147,539,975
-163,330
-329
-9,429
-165.2
-13.4
2030
-289,210,111
-320,120
-625
-18,987
-324.6
-26.2
2040
-374,807,007
-414,844
-800
-22,420
-417.5
-33.9
2050
-470,266,424
-520,496
-1,002
-24,407
-518.5
-42.6
5.4.1  Impact on US and Global GHG Inventory

       As stated in the introduction, climate change is widely viewed as the most significant
long-term threat to the global environment. According to the Intergovernmental Panel on
Climate Change, anthropogenic emissions of greenhouse gases are very likely (90 to 99
percent probability) the cause of most of the observed global warming over the last 50 years.
All mobile sources emitted 31.5 percent of all US GHG in 2006, and have been the fastest -
growing source of US GHG since 1990. Light-duty vehicles are responsible for nearly 60
percent of all mobile source GHGs. For light-duty vehicles, CO2 emissions represent about
95 percent of all greenhouse emissions, and the CO2 emissions measured over the EPA tests
                                       5-25

-------
Draft Regulatory Impact Analysis

used for fuel economy compliance represent over 90 percent of total light-duty vehicle
greenhouse gas emissions.

       This action is an important step towards curbing steady growth of GHG emissions
from cars and light trucks. In the absence of control, GHG emissions worldwide and in the
U.S. are projected to continue steady growth; U.S. GHGs are estimated to make up roughly 15
percent of total worldwide emissions, and the contribution of direct emissions from cars and
light-trucks to this U.S. share is growing overtime, reaching an estimated 20 percent of U.S.
emissions by 2030 in the absence of control.

       As discussed elsewhere in this proposal, this steady rise in GHG emissions is
associated with numerous adverse impacts on human health, food and agriculture, air quality,
and water and forestry resources.
5.5  Non-Greenhouse Gas Emission Inventory

       The reference case emission inventories used for this proposed rule are obtained from
different sources depending on sector.

       For stationary/area sources and aircraft, 2020 projections were used from the 2002
National Emissions Inventory (NEI), Version 3. The  development of these inventories is
documented in the November 27, 2007, memo titled,  "Approach for Developing 2002 and
Future Year National Emission Summaries," from Madeleine Strum to Docket EPA-HQ-
OAR-2007-0491.  That memo summarizes the methodologies and additional reference
documents for criteria air pollutants (CAP) and mobile source air toxics (MSATs).  The
effects of the Clean Air Interstate rule are not included here.

       For onroad mobile sources, the MOVES Draft 2009 model was used that estimates
emissions from light-duty and heavy-duty gasoline and diesel vehicles, except for
motorcycles. For motorcycles, the MOBILE6.2 model was used as run using the NMIM
platform that applies county specific fuel properties and temperatures. For the MOVES
model runs the Vehicle Miles Traveled (VMT) of light-duty gasoline vehicles was adjusted to
account for factors related to this proposal, such as the ten percent rebound effect as described
above in Sections 5.3.3.1.1 and 5.3.3.1.2.

       Most nonroad equipment was modeled with NONROAD2005d using NMIM, which is
a version of the NONROAD that includes the benefits of the two nonroad regulations
published in 2008 (the  locomotive and marine diesel rule and the small spark-ignition and
recreational marine engine rule).35'36

       Inventories for  locomotives and commercial marine vessels are not covered by the
NONROAD model, and they have been updated since the 2002 NEI and its future year
projections were completed.  Thus the  more recent inventory projections published in the
regulatory impact analyses of their respective recent rulemakings were used.35'37
Locomotives and C1/C2 commercial marine vessel inventories come from the spring 2008


                                       5-26

-------
                                                                    Emissions Impacts
final rule, and the C3 commercial marine emission inventory is from the base case inventories
in the June 2009 proposed rule.

       Table 5-25 and Table 5-26 show the total 2020 and 2030 mobile and non-mobile
source inventory projections that were used as the  reference case against which impacts of the
rule were applied.  The impacts, expressed as percentages, are presented below in Sections
5.5.1 through 5.5.3.

        Table 5-25. 2020 Reference Case Emissions by Sector (annual short tons)

Onroad Gasoline
Onroad Diesel
Nonroad SIa
Other Nonroadb
Stationary/Area
Total
voc
1,973,180
129,321
1,289,918
234,870
8,740,057
12,367,346
CO
29,211,716
260,238
14,286,250
1,424,643
11,049,239
56,232,087
NOX
1,934,488
1,353,773
242,828
3,389,761
5,773,927
12,694,778
PM10
96,380
32,733
53,092
230,553
3,194,610
3,607,368
PM2.5
58,990
40,071
49,019
210,509
3,047,714
3,406,303
S02
30,922
4,218
15,413
943,226
7,864,681
8,858,459
TABLE 5-25
CONTINUED
Onroad Gasoline
Onroad Diesel
Nonroad SI3
Other Nonroadb
Stationary/Area
Total
BENZENE
60,742
1,571
36,862
3,760
111,337
214,273
1,3-
BUTADIENE
7,518
830
5,895
929
1,847
17,019
ACETAL-
DEHYDE
14,604
3,743
4,768
9,542
13,118
45,777
FORMAL-
DEHYDE
18,716
10,010
10,240
22,324
23,846
85,136
ACROLEIN
903
475
584
1,013
3,412
6,387
       a Nonroad gasoline, LPG, and CNG engines plus portable fuel containers
       3 Nonroad diesel engines and all locomotive, aircraft, and commercial marine
                                         5-27

-------
Draft Regulatory Impact Analysis
       Table 5-26. 2030 Reference Case Emissions by Sector (annual short tons)

Onroad Gasoline
Onroad Diesel
Nonroad Sf
Other Nonroad"
Stationary/Area
Total
voc
1,800,856
140,959
1,198,679
238,652
8,740,057
12,119,203
CO
32,038,635
219,594
15,815,805
1,411,393
11,049,239
60,534,666
NOX
1,504,390
1,120,656
243,515
3,427,832
5,773,927
12,070,321
PM10
110,796
34,746
55,011
253,572
3,194,610
3,648,735
PM2.5
67,416
26,498
50,816
229,183
3,047,714
3,421,628
SO2
36,011
5,478
17,270
1,426,994
7,864,681
9,350,433
Table 5-26
continued
Onroad Gasoline
Onroad Diesel
Nonroad SI3
Other Nonroadb
Stationary/Area
Total
Benzene
55,692
1,706
39,871
3,764
111,337
212,371
1,3-
Butadiene
6,840
915
6,279
979
1,847
16,859
Acetal-
dehyde
13,354
4,050
5,118
9,579
13,118
45,220
Formal-
dehyde
17,071
10,903
11,229
22,487
23,846
85,536
Acrolein
812
517
629
1,055
3,412
6,425
       a Nonroad gasoline, LPG, and CNG engines plus portable fuel containers
       b Nonroad diesel engines and all locomotive, aircraft, and commercial marine
5.5.1 Downstream Impacts of Program

       As described in the introduction, downstream inventories were generated using
algorithms from EPA's Optimization Model for reducing Emissions of Greenhouse gases
from Automobiles (OMEGA). The OMEGA benefits post-processor produces a national
scale analysis of the benefits (emission reductions, monetized co-benefits) of the analyzed
program.  The non-GHG emission results shown here (Table 5-27) were calculated in a
spreadsheet analysis using algorithms from the Draft MOVES 2009 emission database and
algorithms from the OMEGA post-processor.
                                        5-28

-------
                                                                 Emissions Impacts
           Table 5-27. Downstream Emission Changes of Proposed Program
POLLUTANT
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Paniculate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR YEAR
2020
Short Tons
13.2
25
1.29
80
77,823
32
5,354
218
-2,503
1,698
Percent
Change in
US Total
0.078%
0.054%
0.020%
0.037%
0.138%
0.037%
0.042%
0.006%
-0.028%
0.014%
CALENDAR YEAR
2030
Short Tons
40.2
75
3.89
242
241,939
96
15,560
572
-4,906
5,494
Percent
Change in
US Total
0.238%
0.167%
0.061%
0.114%
0.400%
0.113%
0.129%
0.017%
-0.052%
0.045%
5.5.2 Upstream Impacts of Program

       Fuel production and distribution emission impacts of the proposed program were
estimated in conjunction with the development of life cycle GHG emission impacts, and the
GHG emission inventories discussed above.  The basic calculation is a function of fuel
volumes in the analysis year and the emission factors associated with each process or
subprocess.

       In general this life cycle analysis uses the same methodology as the Renewable Fuel
Standard (RFS2) proposed rule. It relies partially on the "Greenhouse Gases, Regulated
Emissions, and Energy Use in Transportation" (GREET) model, developed by the Department
of Energy's Argonne National Laboratory (ANL), but takes advantage of additional
information and models to significantly strengthen and expand on the GREET analysis.

       Updates and enhancements to the GREET model assumptions include updated crude
oil and gasoline transport emission factors that account for recent EPA emission standards and
modeling,  such as the Tier 4 diesel truck standards published in 2001 and the locomotive and
commercial marine standards finalized in 2008. In addition, GREET does not include air
toxics.  Thus emission factors for the following air toxics were added: benzene, 1,3-
butadiene, formaldehyde, acetaldehyde, and acrolein. These upstream toxics emission factors
were calculated from the 2002 National Emissions Inventory (NEI), a risk and technology
review for petroleum refineries, speciated emission profiles in EPA's SPECIATE database, or
                                       5-29

-------
Draft Regulatory Impact Analysis
the Mobile Source Air Toxics rule (MSAT) inventory for benzene; these pollutant tons were
divided by refinery energy use or gasoline distribution quantities published by the DOE
Energy Information Administration (EIA) to get emission factors in terms of grams per
million BTU of finished gasoline.  The resulting emission factors are presented in Chapter 4
of the draft joint TSD for today's proposed rule.

       Results of these emission inventory impact calculations relative to the reference case
for 2020 and 2030 are shown in Table 5-28 for the criteria pollutants and individual air toxic
pollutants.

       The proposed program is projected to provide reductions in all pollutants associated
with gasoline production and distribution as the projected fuel savings reduce the quantity of
gasoline needed.

            Table 5-28.  Upstream Emission Changes of Proposed Program
POLLUTANT
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Paniculate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR YEAR
2020
Short Tons
-1.8
-8
-1
-163
-7,209
-60
-22,560
-3,075
-13,804
-75,437
Percent
Change in
US Total
-0.010%
-0.017%
-0.017%
-0.076%
-0.013%
-0.071%
-0.178%
-0.090%
-0.156%
-0.610%
CALENDAR YEAR
2030
Short
Tons
-3.4
-15
-2
-320
-14,107
-112
-43,286
-6,003
-27,060
-147,841
Percent
Change in
US Total
-0.020%
-0.033%
-0.032%
-0.151%
-0.023%
-0.131%
-0.359%
-0.175%
-0.289%
-1.220%
5.5.3 Total Program Impact

      Table 5-29 shows the combined impacts of downstream and upstream aspects of the
proposed program.  The fuel production and distribution impacts of the proposed program on
VOC, NOx, PM2.5, and SOx are mainly due to reductions in emissions associated with
gasoline production and distribution as the projected fuel savings of the program reduce the
quantity of gasoline needed. Increases in CO are driven by the rebound effect on VMT,
which are only partially offset by upstream reductions.
      Air toxic emission impacts depend on the relative reductions from upstream emissions

                                        5-30

-------
                                                                  Emissions Impacts

versus increases due to rebound on the downstream emissions.  Relative to 2030 US total
reference case emissions, formaldehyde and benzene emissions are projected to decrease by
0.02 to 0.04 percent, but 1,3-butadiene, acetaldehyde, and acrolein emissions would increase
by 0.03 to 0.22 percent.
         Table 5-29.  Total Non-GHG Emission Changes of Proposed Program
POLLUTANT
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Paniculate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR YEAR
2020
Short Tons
11.5
17
0.2
-84
70,614
-28
-17,206
-2,856
-16,307
-73,739
Percent
Change in
US Total
0.07%
0.04%
0.00%
-0.04%
0.13%
-0.03%
-0.14%
-0.08%
-0.18%
-0.60%
CALENDAR YEAR
2030
Short
Tons
36.8
61
1.8
-77
227,832
-16
-27,726
-5,431
-31,965
-142,347
Percent
Change in
US Total
0.22%
0.13%
0.03%
-0.04%
0.38%
-0.02%
-0.23%
-0.16%
-0.34%
-1.17%
5.6 Model Year Lifetime Analyses

5.6.1 Methodology

       EPA also conducted a separate analysis of the total benefits over the model year
lifetime of 2012 through 2016 model year vehicles. In contrast to the calendar year analysis,
the model year lifetime analysis shows the lifetime impacts of the program on each MY fleet
over the course of its existence.

       In this analysis, a simplified VMT schedule is used.  Rather than using a MY specific
VMT schedule for each MY, a single VMT schedule is used for all five model years.  This
VMT schedule is more fully described in the draft joint TSD. In brief, it was derived using
the same methodology as the MY-specific VMT schedules and is the average of the VMT
schedules from 2012-2030 (Table 5-30).

       All other inputs, including sales, emission factors and achieved emission levels are the
same between the two  analyses.
                                        5-31

-------
Draft Regulatory Impact Analysis
             Table 5-30 - Updated Survival Fraction and Mileage Accumulation by Age
AGE
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
PA
SURVIVAL
FRACTION
0.9950
0.9900
0.9831
0.9731
0.9593
0.9413
0.9188
0.8918
0.8604
0.8252
0.7866
0.7170
0.6125
0.5094
0.4142
0.3308
0.2604
0.2028
0.1565
0.1200
0.0916
0.0696
0.0527
0.0399
0.0301
0.0227
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
PA
MILEAGE
16,932
16,603
16,257
15,814
15,414
14,993
14,545
14,105
13,624
13,192
12,668
12,222
11,705
11,191
10,727
10,283
9,878
9,482
9,090
8,691
8,366
8,126
8,003
7,774
7,587
7,424
7,334
7,200
7,103
7,044
7,042
7,039
7,033
7,021
7,007
6,988
NPA
SURVIVAL
FRACTION
0.9950
0.9741
0.9603
0.9420
0.9190
0.8913
0.8590
0.8226
0.7827
0.7401
0.6956
0.6501
0.6042
0.5517
0.5009
0.4522
0.4062
0.3633
0.3236
0.2873
0.2542
0.2244
0.1975
0.1735
0.1522
0.1332
0.1165
0.1017
0.0887
0.0773
0.0673
0.0586
0.0509
0.0443
0.0385
0.0334
NPA
MILEAGE
18,847
18,408
18,050
17,575
17,142
16,593
16,095
15,493
14,891
14,336
13,689
13,160
12,554
11,945
11,342
10,822
10,383
9,900
9,433
9,033
8,692
8,499
8,246
8,261
8,066
8,066
8,101
8,098
8,096
8,095
8,093
8,092
8,086
8,080
8,064
8,050
                                        5-32

-------
                                                                Emissions Impacts
5.6.2 Results

      The GHG emission reductions are shown for each model year, as are the co-pollutant
impacts (Table 5-31, Table 5-32).

       Table 5-31 - Lifetime GHG Emissions vs. Reference Case (MMT CO2 EQ)

A Downstream Tailpipe
Emission
A Downstream Indirect
A/C
A Downstream
Direct A/C
A Downstream CH4
A Downstream N2O
Total A Downstream

A Upstream CC>2
A Upstream CH4
AUpstream N2O
Total A Upstream

Total Program A GHG
Emissions

Total Program Fuel
Savings (Billion
Barrels)
MY 20 12
-53.19
-5.33
-6.92
0.00
0.02
-65.41

-14.23
-1.69
-0.07
-15.99

-81.41

0.16
MY 2013
-79.99
-9.04
-11.73
0.01
0.03
-100.71

-21.65
-2.58
-0.11
-24.33

-125.05

0.24
MY
2014
-110.75
-12.88
-16.72
0.01
0.05
-140.29

-30.06
-3.58
-0.15
-33.79

-174.08

0.33
MY
2015
-154.99
-17.87
-23.20
0.01
0.06
-195.99

-42.03
-5.01
-0.21
-47.25

-243.23

0.46
MY
2016
-213.28
-20.28
-26.32
0.02
0.10
-259.76

-56.79
-6.76
-0.28
-63.83

-323.59

0.63
Program
Total
-612.19
-65.39
-84.89
0.05
0.26
-762.17

-164.75
-19.62
-0.82
-185.19

-947.36

1.82
                                       5-33

-------
Draft Regulatory Impact Analysis

      Table 5-32 - Lifetime non-GHG Emissions vs. Reference Case (Short Tons)

Downstream
AVOC
ANOX
APM2.5
AGO
ASO2
A Benzene
A 1,3 Butdiene
A Formaldehyde
A Acetaldehyde
A Acrolein

Upstream
AVOC
ANOX
A PM 2.5
AGO
AS02
A Benzene
A 1,3 Butdiene
A Formaldehyde
A Acetaldehyde
A Acrolein

Total
AVOC
ANOX
APM2.5
AGO
ASO2
A Benzene
A 1,3 Butdiene
A Formaldehyde
A Acetaldehyde
A Acrolein
MY 20 12

1,495
3,991
152
63,783
-1,234
65.1
10.8
25.9
20.3
1.0


-37,188
-10,888
-1,510
-3,549
-6,807
-80.4
-0.9
-28.2
-3.7
-0.5


-35,694
-6,897
-1,358
60,234
-8,041
-15
10
-2
17
1
MY 2013

2,373
6,602
241
102,779
-1,877
103.4
17.1
41.1
32.2
1.7


-56,572
-16,564
-2,297
-5,398
-10,354
-122.3
-1.3
-42.9
-5.7
-0.8


-54,199
-9,962
-2,056
97,381
-12,232
-19
16
-2
27
1
MY 20 14

3,336
9,469
338
145,598
-2,607
145.4
24.1
57.9
45.3
2.3


-78,562
-23,002
-3,190
-7,496
-14,379
-169.9
-1.8
-59.5
-7.9
-1.1


-75,226
-13,533
-2,852
138,101
-16,986
-24
22
-2
37
1
MY 2015

4,751
13,485
482
207,338
-3,645
207.1
34.3
82.4
64.5
3.3


-109,850
-32,163
-4,460
-10,482
-20,106
-237.6
-2.6
-83.2
-11.0
-1.5


-105,099
-18,678
-3,978
196,856
-23,751
-30
32
-1
53
2
MY 2016

6,763
20,082
686.074
295,494
-4,925
294.2
48.8
117.1
91.6
4.7


-148,418
-43,455
-6,026
-14,162
-27,165
-321.0
-3.5
-112.5
-14.9
-2.1


-141,655
-23,373
-5,340
281,332
-32,090
-27
45
5
77
3
Program
Total

18,717
53,629
1,899
814,992
-14,288
815
135
324
254
13


-430,591
-126,072
-17,484
-41,087
-78,812
-931
-10
-326
-43
-6


-411,874
-72,443
-15,585
773,905
-93,099
-116
125
-2
211
7
5.7 Alternative 4% and 6% Scenarios

      For this proposal, two alternative control scenarios were evaluated characterized by
4% and 6% annual growth in the GHG standards from the MY 2011 standard. Other than the
                                      5-34

-------
                                                                  Emissions Impacts
standards, these scenarios share all inputs with the proposed EPA program.  Only GHG
reductions and fuel savings are shown for these programs.

5.7.1 4% Scenario
5.7.1.1
Standards and Achieved Levels
The program standards are shown in Table 5-33 and the achieved levels are shown in Table
5-34.

                            Table 5-33: 4% Scenario Standards
MODEL
YEAR
2012
2013
2014
2015
2016
PA
EMISSION
LEVEL
276
265
255
246
237
NPA EMISSION
LEVEL
366
355
342
331
318
ANTICIPATED
MY EMISSION
LEVEL
310
298
286
275
264
                          Table 5-34: 4% Scenario Achieved Levels
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
275
266
258
249
237
ANTICIPATED
NPA EMISSION
LEVEL
378
368
352
338
319
ANTICIPATED
MY EMISSION
LEVEL
314
304
292
279
264
    5.7.1.2  Results

       Results are shown relative to the same reference scenario as the proposed EPA
program. Both calendar year and model year lifetime results are shown.
                                        5-35

-------
Draft Regulatory Impact Analysis
  Table 5-35 - Downstream CY GHG Reductions and Fuel Savings vs. Reference Case

Downstream
A CO2 excluding indirect A/C controls (MMT CO2 EQ)
AIndirect A/C CO2(MMT CO2 EQ)
A Direct A/C HFC (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
AN2O(MMTCO2EQ)
A Total GHG (MMT CO2 EQ)

Upstream
A CO2 (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
AN2O(MMTCO2EQ)
A Total GHG

Total
A Total GHG
A Fuel Consumption (Annual, Billion gallons)
CY
2020

-97.7
-11.0
-13.5
0
0
-122.2


-26.4
-4.1
-0.1
-30.5


-152.7
-12.2
CY
2030

-196.6
-21.1
-27.2
0
0
-244.8


-52.9
-8.0
-0.3
-61.2


-306.0
-24.5
CY
2040

-255.5
-27.2
-32.1
0
0.1
-314.7


-68.7
-10.4
-0.3
-79.4


-394.4
-31.8
CY
2050

-320.7
-34.1
-34.9
0
0.1
-389.6


-86.2
-13.0
-0.4
-99.6


-489.2
-39.9
                                     5-36

-------
                                                                 Emissions Impacts
         Table 5-36 - Total Model Year Lifetime GHG Reductions vs. Baseline

Downstream
A CO2 excluding indirect A/C controls
(MMT CO2 EQ)
AIndirect A/C CO2(MMT CO2 EQ)
A Direct A/C HFC (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
A N2O (MMT CO2 EQ)
A Total GHG (MMT CO2 EQ)

Upstream
A CO2 (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
A N2O (MMT CO2 EQ)
A Total GHG

Total
A Total GHG
A Fuel Consumption (Annual, Billion
gallons)
MY
2012

-22.1
-5.3
-6.9
0.0
0.0
-34.3


-6.7
-0.8
0.0
-7.5


-41.8
-3.09
MY
2013

-55.2
-9.0
-11.7
0.0
0.0
-75.9


-15.6
-1.9
-0.1
-17.6


-93.5
-7.3
MY
2014

-100.3
-12.9
-16.7
0.0
0.0
-129.9


-27.5
o o
-J.J
-0.1
-30.9


-160.8
-12.7
MY
2015

-145.4
-17.9
-23.2
0.0
0.1
-186.4


-39.7
-4.7
-0.2
-44.6


-231.0
-18.4
MY
2016

-198.8
-20.3
-26.3
0.0
0.1
-245.3


-53.3
-6.3
-0.3
-59.9


-305.2
-24.7
Program
Total

-521.8
-65.4
-84.9
0.0
0.2
-671.8


-142.8
-17.0
-0.7
-160.5


-832.3
-66.1
5.7.2 6% Scenario

    5.7.2.1  Standards and Achieved Levels

       The program standards are shown in Table 5-33 and the achieved levels are shown in
Table 5-34.

                           Table 5-37: 6% Scenario Standards
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
269
255
241
228
216
ANTICIPATED
NPA EMISSION
LEVEL
362
342
323
306
290
ANTICIPATED
MY EMISSION
LEVEL
304
287
270
255
241
                                       5-37

-------
Draft Regulatory Impact Analysis
                         Table 5-38: 6% Scenario Achieved Levels
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
270
258
244
230
216
ANTICIPATED
NPA EMISSION
LEVEL
375
352
332
313
290
ANTICIPATED
MY EMISSION
LEVEL
309
292
275
259
241
    5.7.2.2  Results

       Results are shown relative to the same reference scenario as the proposed EPA
program. Both calendar year and model year lifetime results are shown.

       Table 5-39 -CY GHG Emissions and Fuel Consumption vs. Reference Case

Downstream
A CO2 excluding indirect A/C controls (MMT CO2 EQ)
AIndirect A/C CO2(MMT CO2 EQ)
A Direct A/C HFC (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
A N2O (MMT CO2 EQ)
A Total GHG (MMT CO2 EQ)

Upstream
A CO2 (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
AN2O(MMTCO2EQ)
A Total GHG

Total
A Total GHG
A Fuel Consumption (Annual, Billion gallons)
CY
2020

-146.8
-11.0
-13.5
0.0
0.0
-171.3


-38.4
-5.5
-0.1
-44.0


-215.2
-17.8
CY
2030

-290.8
-21.3
-27.2
0.0
0.0
-339.2


-75.9
-10.7
-0.4
-87.0


-426.2
-35.1
CY
2040

-377.2
-27.5
-32.1
0.0
0.1
-436.6


-98.4
-14.0
-0.4
-112.7


-549.3
-45.5
CY
2050

-473.3
-34.4
-34.9
0.0
0.1
-542.4


-123.4
-17.4
-0.5
-141.4


-683.9
-57.1
                                       5-38

-------
                                                            Emissions Impacts
Table 5-40 -MY Lifetime GHG Emissions and Fuel Consumption vs. Reference Case

Downstream
A CO2 excluding indirect A/C controls
(MMT CO2 EQ)
AIndirect A/C CO2(MMT CO2 EQ)
A Direct A/C HFC (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
A N2O (MMT CO2 EQ)
A Total GHG (MMT CO2 EQ)

Upstream
A CO2 (MMT CO2 EQ)
A CH4 (MMT CO2 EQ)
A N2O (MMT CO2 EQ)
A Total GHG

Total
A Total GHG
A Fuel Consumption (Annual, Billion
gallons)
MY
2012

-36.5
-5.3
-6.9
0.0
0.0
-48.7


-10.2
-1.2
-0.1
-11.4


-60.2
-4.7
MY
2013

-96.8
-9.0
-11.7
0.0
0.0
-117.5


-25.7
-3.1
-0.1
-28.9


-146.4
-11.9
MY
2014

-162.5
-12.9
-16.7
0.0
0.1
-192.0


-42.6
-5.1
-0.2
-47.9


-239.9
-19.7
MY
2015

-225.7
-17.9
-23.2
0.0
0.1
-266.7


-59.2
-7.1
-0.3
-66.6


-333.3
-27.4
MY
2016

-292.9
-20.3
-26.3
0.0
0.1
-339.3


-76.1
-9.1
-0.4
-85.6


-424.9
-35.2
Program
Total

-814.4
-65.4
-84.9
0.1
0.4
-964.2


-213.9
-25.5
-1.1
-240.4


-1204.7
-99.0
                                    5-39

-------
Draft Regulatory Impact Analysis




5.A Appendix to Chapter 5:  Details of the TLAAS Impacts Analysis

5.A.1 Introduction and Summary

       The TLAAS program allows manufacturers with total domestic sales of less than
400,000 vehicles during model year 2009 to place up to 100,000 vehicles from model years
2012-2015 into a separate fleet. This separate fleet is subject to a 25% less stringent standard
than the manufacturer's primary fleet (subject to various further constraints described in
section III of the preamble and in the proposed rule itself).

       Several manufacturer decisions and marketplace events determine the impacts of the
TLAAS program.  This appendix presents a sensitivity analysis that brackets the impact of the
program, and provides additional details on the assumptions made in the EPA emission
analysis.

       Although the bracketing analyses presented here range from 0 to 25 MMT of €62
emissions, in all cases the TLAAS program has a proportionally small impact (< 3%) on the
total program benefits over the model years 2012-2016.

       Under the estimation procedure used in the emission inventory analysis (as opposed to
the bracketing analysis mentioned immediately above), the TLAAS is projected to result in an
approximately 3.4 MMT decrease in greenhouse gas benefits from this rule over the lifetime
of vehicles manufactured in model years 2012-2015 (assuming that it is technically feasible
for all TLA AS-eligible producers to meet the otherwise-applicable GHG standards for those
years, a dubious assumption given the very short lead times available).

5.A.2 Factors Determining the Impact of the TLAAS

       The greatest challenge to accurately estimating the impacts of the TLAAS  are
uncertainties about manufacturer eligibility and manufacturer usage of the program. There is
a third, albeit smaller uncertainty, concerning the size of the vehicles placed in the program.

Eligibility

       Up to eleven major manufacturers are potentially eligible for TLAAS  based on
preliminary EPA analysis of projected domestic sales for model year 2009. These
manufacturers are  Porsche, Tata, Mazda, Mitsubishi, Suzuki, Daimler, Subaru, BMW,
Volkswagen, Hyundai, and Kia.

       Manufacturers such as Hyundai, Kia, Mazda, and Volkswagen are preliminarily
estimated at 2009 domestic sales bordering 400,000. If none of these four manufacturers are
eligible for the TLAAS program, the program covers up to 700,000 vehicles.  If all four are
included, the program increases in size by approximately 50% to 1.1 million vehicles.
                                        5-40

-------
                                                                 Emissions Impacts

       The impacts of the program therefore partially depend on manufacturer eligibility.

Manufacturer Usage

       By reducing the compliance burden, the TLAAS provides needed lead time flexibility
to manufacturers in order to comply with the Light Duty Vehicle Greenhouse Gas Program in
the short term, and provides needed lead time for these manufacturers to bring their entire
fleet into compliance with the stringent 2016 MY standards.  However, it is unclear whether
manufacturers will participate in the TLAAS program to the fullest extent allowed, as there
are two disincentives to fully utilizing the TLAAS.

       Vehicles in the TLAAS fleet may consume more fuel than comparably sized vehicles
in the primary fleet. Assuming consumers place some weight on fuel economy when
purchasing a vehicle, manufacturers with TLAAS fleets may thus place their vehicles at a
competitive disadvantage in the marketplace.

       Further, at the cessation of the TLAAS program, manufacturers will need their fleet to
meet the more stringent main program standards. If a manufacturer takes full advantage of
the program by using the maximum 25% additional emission allotment, they may place
themselves at a technological disadvantage when the program ends. Both in terms of
engineering and manufacturing, a manufacturer is unlikely to want to  fall behind its
competitors. To avoid this scenario, a manufacturer may make gradual gains over the
TLAAS program, and gradually use less of the 25% additional emission allotment.

       Because of these disincentives, manufacturers may likewise choose to not fully utilize
the TLAAS vehicle production volumes.

Size and Classification of the Vehicles Placed in the TLAAS Fleet

       As the TLAAS program allows 25% additional emissions over the footprint-based
main fleet standards, the size of the vehicles placed in the TLAAS fleet is significant. If a
manufacturer places small but high emitting vehicles in the TLAAS fleet (ie, Porsche
Carrera), the impact of the program is less than if large and high emitting vehicles are placed
in the TLAAS fleet.

       A manufacturer who utilized the TLAAS fleet for small vehicles would necessarily
have a proportionally lower net impact.  Similarly, due to the two distinct footprint curves, the
choice whether to place cars or trucks in the TLAAS fleet will also determine impact.

5.A.3 Bounding Analysis of TLAAS Impact

       This section provides upper and lower bounds for the potential impacts from the
TLAAS, and then describes the inputs used in the emission analysis.
       TLAAS is an optional program which can be used for a limited number of eligible
vehicles to achieve compliance with the Light Duty Vehicle Greenhouse Gas Program.

                                        5-41

-------
Draft Regulatory Impact Analysis

Consequently, no manufacturer is obligated to use the program, and the lower bound of the
program impact could theoretically be zero. This is considered a highly unlikely scenario, as
several manufacturers are anticipated to use the TLAAS to meet their compliance targets
given the lack of lead time for these manufacturers to make the major conversions necessary
to meet the standards.

       Conversely, as an upper bound, every manufacturer could use their full allocation on
their largest vehicle, could potentially increase sales of those vehicles to  100,000 over the four
year period, and could use the full 25% "cushion" for each of these vehicles. This is also an
unlikely scenario, as it would require companies such as Porsche and BMW to sell specific
vehicle models (such as the Porsche Boxster, or the Rolls Royce Phantom) in unprecedented
numbers.

       As a boundary analysis, EPA analyzed these upper and lower bound scenarios . The
GHG savings from the lower bound program was estimated at 950 MMT GHG reduced over
lifetime of model years 2012-2016 (i.e. impact of the TLAAS is  zero), while the upper bound
impact was 925 MMT GHG reduced over the same period.  Thus, the maximum potential
impact of the program, even under this most extreme scenario is approximately 25 MMT.

       As noted, neither of these scenarios is remotely likely.  However, the point of the
bounding analysis is to show that the greatest possible impact of the proposed TLAAS is still
relatively minimal.

5.A.4 Approach used for Estimating TLAAS Impact

       Having bounded the analysis, a third approach was used for the emission modeling
described in DRIA chapter 5. In this analysis, all eleven manufacturers were assumed to use
the default vehicle allocation schedule from the TLAAS. This is a conservative estimate, as
several of the manufacturers are unlikely to utilize their allocation due to either lack of need,
or the disincentives discussed above.
                    Table 5-41:TLAAS Default Vehicle Production Volumes
MODEL
YEAR
Sales Volume
2012
40,000
2013
30,000
2014
20,000
2015
10,000
                                        5-42

-------
                                                                  Emissions Impacts
       The allocation was split evenly between cars and tracks for each manufacturer. The
TLAAS fleet was assumed to emit as much CO2 per mile as expected from the largest
footprint car or track in each manufacturer's fleet. This estimate combines the impact of the
25% additional emission allotment and the vehicle size factors discussed above. These
vehicles were then proportionally averaged into the manufacturer's GHG score. This resulted
in an emission impact of approximately 3.4 MMT CCh over the lifetime of the 2012 -2015
MY vehicles.

       The gram per mile impacts are  listed here for each of these scenarios.

Model Year
2012
2013
2014
2015
2016
TLAAS impact (Grams CO2
Emissions Per Mile)
Lower
Bound
Scenario
0
0
0
0
0
Upper
Bound
Scenario
2.0
1.5
1.0
0.5
0
Estimate
Used In
Emission
Analysis
0.3
0.2
0.2
0.1
0.0
                                        5-43

-------
Draft Regulatory Impact Analysis

        References

        All references can be found in the EPA DOCKET:  EPA-HQ-OAR-2009-0472.
1 Argonne National Laboratory. The Greenhouse Gases, Regulated Emissions, and Energy Use in
Transportation (GREET) Model versions 1.7 and 1.8.
http://www.transportation.anl.gov/modeling_simulation/GREET/

2 Mcculloch A.; Lindley A. A. From mine to refrigeration: a life cycle inventory analysis of the production of
HFC-134a .. ; International journal of refrigeration  2003, vol. 26, no8, pp. 865-872

3 Intergovernmental Panel on Climate Change.  Chapter 2.  Changes in Atmospheric Constituents and in
Radiative Forcing. September 2007. http://www.ipcc.ch/pdf/assessment-report/ar4/wgl/ar4-wgl-chapter2.pdf

4  EPA. Draft MOVES 2009. http://www.epa.gov/otaq/models/moves/index.htm

5 U.S. EPA 2009. Updated OMEGA Post-Processor Spreadsheet. August 15, 2009.

6 Updated Tables to MOVES in Docket. Samplevehiclepopulation and emissionbyage.

7 John Koupal, Richard Rykowski, Todd Sherwood, Ed Nam. "Documentation of Updated Light-duty Vehicle
GHG Scenarios." Memo to Docket ID No. EPA-HQ-OAR-2008-0318

8 MOVES documentation and technical documents can be seen at
http://www.epa.gov/otaq/models/moves/index.htm.

9 Reference OMEGA Peer Review

10 U.S. EPA. Draft Regulatory Impact Analysis: Changes to Renewable Fuel Standard Program.  Chapters 2
and 3.May 26, 2009.

11 Argonne National Laboratory.  The Greenhouse Gases, Regulated Emissions, and Energy Use in
Transportation (GREET) Model versions 1.7 and 1.8.
http://www.transportation.anl. gov/modeling_simulation/GREET/

12 U.S. EPA. 2008. RFS2 Modified version of GREET1.7 Upstream Emissions Spreadsheet, October 31, 2008.
13
  U.S. EPA, Achieved CO2 standards worksheet. 2009.
14 NHTSA.  2009. Average Fuel Economy Standards, Passenger Cars and Light Trucks, Model Year 2011.
Docket ID:  NHTSA-2009-0062-0001.
http://www.nhtsa.dot.gov/portal/nhtsa_static_file_downloader.jsp?file=/staticfiles/DOT/NHTSA/Rulemaking/Ru
les/Associated Files/CAFE_Updated_Final_Rule_MY2011 .pdf

  NHTSA. 2009. Average Fuel Economy Standards, Passenger Cars and Light Trucks, Model Year 2011.
http://www.nhtsa.dot.gov/portal/nhtsa_static_file_downloader.jsp?file=/staticfiles/DOT/NHTSA/Rulemaking/Ru
les/Associated Files/CAFE_Updated_Final_Rule_MY2011 .pdf

16U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1.  August 2009..

17 Energy Information Administration. Annual Energy Outlook 2009. Supplemental Transportation Tables.
April 2009. http://www.eia.doe.gov/oiafaeo/supplement/sup_tran.xls
                                              5-44

-------
                                                                             Emissions Impacts
18 NHTSA.  2009.  Average Fuel Economy Standards, Passenger Cars and Light Trucks, Model Year 2011.
Docket ID:  NHTSA-2009-0062-0001.
http://www.nhtsa.dot.gov/po rtal/nhtsa_static_file_downloader.jsp?file=/staticfiles/DOT/NHTSA/Rulemaking/Ru
les/Associated Files/CAFE_Updated_Final_Rule_MY2011 .pdf

19 NHTSA.  2009.  Average Fuel Economy Standards, Passenger Cars and Light Trucks, Model Year 2011.
Docket ID:  NHTSA-2009-0062-0001.
http://www.nhtsa.dot.gov/po rtal/nhtsa_static_file_downloader.jsp?file=/staticfiles/DOT/NHTSA/Rulemaking/Ru
les/Associated Files/CAFE_Updated_Final_Rule_MY2011 .pdf

20 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. August 2009..

21 EPA.  Emission Facts: Average Carbon Dioxide Emissions Resulting from Gasoline and Diesel Fuel.
EPA420-F-05-001  February 2005

22 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. August 2009..

        23 Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy. Transportation
Energy Data Book: Edition 27. Chapter 4. 2008.

24 NHTSA.  Vehicle Survivability and Travel Mileage Schedules.  2006.

25 Draft TSD Chapter 4

26 NHTSA.  Vehicle Survivability and Travel Mileage Schedules.  2006.

27 Draft TSD Chapter 4

28 EPA.  Final Technical Support Document.Fuel Economy Labeling of Motor Vehicle Revisions to Improve
Calculation of Fuel Economy Estimates

29 John Koupal, Richard Rykowski, Todd Sherwood, Ed Nam.  "Documentation of Updated Light-duty Vehicle
GHG Scenarios." Memo to Docket ID No. EPA-HQ-OAR-2008-0318

30 Draft RIA chapter 2.

31 Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy. Transportation Energy Data
Book: Edition 27. 2008.

32 EPA.  Final Technical Support Document. Fuel Economy Labeling of Motor Vehicle Revisions to Improve
Calculation of Fuel Economy Estimates

33 Average Carbon Dioxide Emissions Resulting from Gasoline and Diesel Fuel.  EPA420-F-05-001 February
2005. http://www.epa.gov/otaq/climate/420f05001 .htm

34 Craig Harvey, EPA, "Calculation of Upstream Emissions for the GHG Vehicle Rule." 2009.

35 Control of Emissions of Air Pollution From Locomotive Engines and Marine Compression-Ignition Engines
Less Than 30 Liters per Cylinder, Republication, Final Rule (Federal Register Vol 73, No. 126, page 37096,
June 30, 2008).

36 Control of Emissions From Nonroad Spark-Ignition Engines and Equipment, Final Rule (Federal Register Vol
73, No. 196, page 59034, October 8, 2008).


                                              5-45

-------
Draft Regulatory Impact Analysis
37 Draft Regulatory Impact Analysis: Control of Emissions of Air Pollution from Category 3 Marine Diesel
Engines, Chapter 3. This is available in Docket OAR-2007-0121at http://www.regulations.gov/
                                              5-46

-------
                           Vehicle Program Costs Including Fuel Consumption Impacts

 CHAPTER 6: Vehicle Program Costs Including Fuel
                   Consumption Impacts

       This chapter presents the costs of the proposed GHG vehicle program including the
costs associated with addition of new technology and savings associated with improved fuel
consumption. In section 6.1, vehicle compliance costs are presented on a per-car and per-
truck basis for each manufacturer and the industry as a whole. Vehicle compliance costs are
also presented on an annual basis for each manufacturer and the industry as a whole.  Where
appropriate, net present values are presented at both a 3 percent and a 7 percent discount rate
for annual costs in the years 2012 through 2050. In section 6.2, the cost per ton of GHG
reduced is presented as a result of the proposal. In section 6.3, fuel consumption impacts are
presented on a per-year basis for cars and trucks in terms of gallons saved and in terms of
dollars saved. In section 6.4, the vehicle program costs and fuel consumption impacts are
summarized.  This chapter does not present costs associated with noise, congestion, accidents
and other economic impacts associated with increased driving that could result from the
proposed program. Such impacts are presented in Chapter 8 of this draft RIA.

 6.1 Vehicle Program Costs

       Chapter 4 of this draft RIA presents the outputs of the OMEGA model for the model
year 2016.  Here, we build on those results and calculate estimated costs for each model year
beginning with 2012 and going through 2050. We do this both on a per-vehicle basis and an
annual basis.  Costs here include costs associated with the proposed A/C credit program. For
details on the individual technology costs please refer to Chapters of the draft joint TSD. For
details on the OMEGA model inputs (i.e., how the individual technology costs are combined
into package costs) please refer to Chapter 1 of this draft RIA. For  details on the A/C costs,
please refer to Chapter 2 of this draft RIA.

6.1.1 Vehicle Compliance Costs on a Per-Vehicle Basis

       As stated above, Chapter 4 of this draft RIA presents the cost per vehicle for each
manufacturer in the 2016 model year. Those 2016 MY costs are reproduced in Table 6-1. To
estimate the cost per vehicle for model years 2012 through 2015, we first looked at the
projected CO2 levels for each manufacturer's fleet for each year 2011 though 2016.  Those
CO2 levels are presented in Table 6-2 for cars and Table 6-3 for trucks.A  The achieved CO2
levels for 2012-2015 were  derived using the same process described in chapter 5 of the DRIA.
Starting with the calculated manufacturer, vehicle class, and model  year specific achieved
standards, we estimated the cost effective environmentally neutral credit trading based on the
ANote that the 2012-2015 CO2 levels are estimates based upon assumptions of manufacturer fleetwide CO2
averages in 2011, which are extrapolated from a 2008 base fleet.  Consequently, the average CO2 emission
levels for some manufacturers are potentially too high for the 2011 MY which makes the transition to the
2012MY appear as a more significant change. As a result, 2012MY costs represent a large percentage of the
total costs. As an example, the 2012MY cost for Suzuki as shown in Table 6-5 is approximately 60% of the
2016MY cost. In reality, the transition between MY 2011 and MY 2016 may be significantly smoother, and is
likely to be smoother due to multiyear planning.
                                         6-1

-------
Regulatory Impact Analysis
2016 achieved levels predicted by OMEGA.  Based on this process, we projected the most
likely achieved levels for 2012-2015 for each manufacturer.

       There are some differences between cost effective achieved levels and the achieved
levels shown in DRIA chapter 5. As shown here, the cost effective achieved levels for the
intermediate years were derived in the following manner. MY 2011 baseline CO2 was
determined from the reference fleet file.1 MY 2016 achieved CCh was determined from the
OMEGA output described in DRIA chapter 4. To determine the intermediate years, an
interpolation was performed between these two points. Two different forms of interpolation
were used. For manufacturers that fully comply with the 2016 standards, the change between
2011 and 2016 was weighted by the percent change in their achieved target for each year (as
determined in DRIA chapter 5). For the manufacturers that do not fully comply in 2016,
these manufacturers improved by 20% of their total change each year.

       Two manufacturers, Subaru and Mitsubishi, had their improvement front loaded in
order to produce early year compliance. These companies are anticipated to comply with the
intermediate year standards, but the 2008 base fleet may understate their expected
performance.8 The analysis behind the cost effective achieved levels is contained in the EPA
docket.2

       We then used these CO2 values to generate ratios that could be applied to the 2016
MY costs to arrive at cost estimates for each of the intervening years. This methodology is
based, in part, on the credit carry-forward and carry-back provisions contained in the
proposal.  However, we must also remember  that the technology costs and, subsequently, the
package costs in the 2016 MY have undergone some learning effects as described in Chapter
3 of the draft joint TSD.   We compared the 2016 MY package costs to each of the intervening
years and the results, on a percentage basis, are shown in Table 6-4. We have also done this
for the years following 2016 to reflect the effects of the near term and long term ICMs as
described in Chapter 3 of the draft joint TSD. The process for estimating costs in the
intervening years is best understood by way of an example: General Motors cars are
estimated to incur a cost of $969 in the 2016 MY while achieving  a CO2 average of 240 g/mi;
for the 2011 and 2012 MYs, GM cars are projected to achieve a CO2  average of 306 and 276,
respectively.  We can apply the ratio (306-276)/(306-240)=0.45 to GM's 2016 cost of $969,
and then apply the 2012 relative to 2016 cost factor of 117%, to arrive at an estimated 2012
cost of $507.c We then carry out this process for each manufacturer for each year to arrive at
the results presented in Table 6-5 for cars, Table 6-6 for trucks, and Table 6-7 for cars and
trucks combined.  Table 6-8 shows the industry average cost per car, cost per truck, and cost
per vehicle (car/truck combined) for the 2012 and later model years.D
EIbid

c Numbers in the text are rounded for clarity so results using numbers shown in the text may not match those in
tables.

D Note that the costs per car, truck and vehicle presented here do not include possible maintenance savings
associated with the new A/C systems.  They also do not include maintenance costs associated with low friction
lubes and low rolling resistenance tires. We include higher new vehicle costs for these latter items but do not
                                         6-2

-------
                             Vehicle Program Costs Including Fuel Consumption Impacts
           Table 6-1 Cost per Car and Truck, including A/C, for the 2016 MY (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
$/CAR
$1,700
$1,331
$1,630
$1,434
$969
$606
$739
$741
$946
$1,067
$1,012
$1,548
$902
$1,093
$1,269
$599
$1,626
$968
$/TRUCK
$1,664
$1,505
$1,356
$1,485
$1,781
$695
$1,679
$1,177
$1,029
$1,263
$1,193
$666
$1,328
$1,263
$673
$435
$949
$1,213
              Table 6-2 Projected CO2 Levels for MYs 2011-2016, Cars Only (g/mi CO2)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
201 1MY
308.0
313.1
308.6
311.0
305.8
263.2
282.2
280.5
286.8
287.1
278.4
339.1
304.0
284.5
347.3
250.9
291.5
285.1
2012MY
292.5
279.0
293.1
276.5
276.4
251.2
260.0
263.0
260.1
258.1
262.7
317.9
272.9
247.0
328.9
240.5
278.3
264.0
2013MY
278.2
270.2
278.8
268.3
267.7
240.3
251.8
254.6
251.8
246.5
255.1
297.9
258.6
238.4
311.8
231.3
266.4
254.5
2014MY
263.3
261.3
263.9
259.5
259.1
228.9
243.5
245.8
243.0
231.0
246.9
277.3
249.9
229.7
294.1
221.5
253.8
244 .4
2015MY
248.4
250.0
249.0
246.2
246.0
217.4
232.3
234.4
231.4
215.6
235.8
256.7
241.3
217.9
276.4
211.7
241.2
232.4
2016MY
235.3
219.6
236.0
226.4
239.8
207.7
226.8
223.1
221.0
211.7
217.4
237.9
231.3
203.3
260.4
203.7
230.4
220.9
account for higher replacement costs during vehicle lifetimes even though oil is changed many times and tires
are changed once or twice. We intend to include these maintenance costs and savings in our final rule analysis.
                                             6-3

-------
Regulatory Impact Analysis
            Table 6-3 Projected CO2 Levels for MYs 2011-2016, Trucks Only (g/mi CO2)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
201 1MY
357.3
386.7
423.0
413.4
410.5
346.7
357.6
353.9
332.7
337.2
387.3
369.6
334.9
356.7
362.5
353.4
417.6
383.6
2012MY
341.2
365.7
397.6
382.1
392.7
338.3
330.6
339.2
326.0
327.2
370.3
361.0
294.4
314.7
355.2
345.8
398.9
365.3
2013MY
326.5
354.6
373.7
370.3
380.8
331.2
320.3
329.1
314.9
322.4
358.9
353.7
275.9
306.1
349.3
339.6
381.5
355.5
2014MY
311.2
344.0
349.0
360.2
369.3
323.5
308.4
317.8
302.7
317.9
348.7
345.8
264.9
296.7
342.7
332.8
363.4
345.8
2015MY
295.8
326.5
324.3
342.5
351.0
315.8
293.1
302.8
287.4
311.5
331.1
337.9
254.0
283.8
336.1
325.9
345.3
331.7
2016MY
282.4
314.1
301.7
316.6
308.4
310.0
257.6
291.4
267.9
307.0
321.5
332.0
240.9
280.7
331.5
321.0
329.2
310.9
           Table 6-4 Package Costs Measured Relative to the Package Costs for the 2016MY
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022+
PACKAGE COSTS
RELATIVE TO 20 16
117%
115%
108%
102%
100%
100%
100%
100%
100%
100%
92%
                                            6-4

-------
              Vehicle Program Costs Including Fuel Consumption Impacts
 Table 6-5 Cost per Car, including A/C, by Manufacturer (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2012MY
$426
$571
$408
$686
$507
$155
$348
$265
$451
$483
$306
$381
$454
$593
$315
$155
$412
$374
2013MY
$801
$703
$768
$831
$643
$288
$466
$385
$578
$661
$446
$725
$648
$713
$596
$286
$770
$531
2014MY
$1,130
$797
$1,084
$945
$740
$406
$558
$484
$681
$858
$566
$1,022
$726
$798
$841
$404
$1,086
$663
2015MY
$1,425
$918
$1,367
$1,123
$897
$512
$680
$608
$814
$1,035
$724
$1,289
$796
$916
$1,060
$509
$1,369
$813
2016MY
$1,700
$1,331
$1,630
$1,434
$969
$606
$739
$741
$946
$1,067
$1,012
$1,548
$902
$1,093
$1,269
$599
$1,626
$968
Table 6-6 Cost per Truck, including A/C, by Manufacturer (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2012MY
$420
$510
$333
$564
$364
$188
$532
$325
$125
$488
$363
$179
$673
$820
$186
$120
$237
$358
2013MY
$786
$763
$634
$760
$596
$337
$720
$536
$326
$709
$592
$323
$959
$967
$330
$213
$446
$539
2014MY
$1,109
$956
$895
$882
$777
$475
$893
$735
$515
$874
$758
$455
$1,069
$1,079
$465
$300
$629
$682
2015MY
$1,398
$1,276
$1,128
$1,112
$1,061
$599
$1,107
$984
$736
$1,101
$1,042
$574
$1,169
$1,238
$586
$378
$794
$886
2016MY
$1,664
$1,505
$1,356
$1,485
$1,781
$695
$1,679
$1,177
$1,029
$1,263
$1,193
$666
$1,328
$1,263
$673
$435
$949
$1,213
                             6-5

-------
Regulatory Impact Analysis
    Table 6-7 Cost per Vehicle (car/truck combined), including A/C, by Manufacturer (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
GeneralMotors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2012MY
$424
$534
$376
$641
$443
$167
$384
$304
$384
$484
$328
$354
$527
$683
$281
$142
$376
$368
2013MY
$797
$738
$710
$806
$622
$305
$516
$473
$529
$670
$501
$661
$747
$806
$528
$260
$712
$534
2014MY
$1,124
$887
$998
$922
$756
$428
$623
$635
$651
$861
$639
$944
$831
$892
$738
$368
$1,003
$670
2015MY
$1,418
$1,080
$1,259
$1,119
$968
$538
$759
$813
$801
$1,045
$829
$1,201
$905
$1,019
$935
$465
$1,265
$838
2016MY
$1,691
$1,408
$1,509
$1,452
$1,311
$632
$907
$973
$959
$1,095
$1,070
$1,444
$1,023
$1,147
$1,117
$545
$1,508
$1,050
        Table 6-8 Industry Average Cost per Car, Truck, and Combined by Year (2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
$/CAR
$374
$531
$663
$813
$968
$968
$968
$968
$968
$968
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$/TRUCK
$358
$539
$682
$886
$1,213
$1,213
$1,213
$1,213
$1,213
$1,213
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$/VEHICLE
$368
$534
$670
$838
$1,050
$1,047
$1,044
$1,042
$1,040
$1,039
$955
$955
$955
$955
$954
$954
$954
$953
$953
$953
$953
$953
$953
$953
                                             6-6

-------
                           Vehicle Program Costs Including Fuel Consumption Impacts
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$890
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$1,116
$953
$953
$953
$953
$953
$953
$953
$953
$953
$953
$953
$953
$953
$953
$953
6.1.2  Vehicle Compliance Costs on a Per-Year Basis

       Given the cost per car and cost per truck estimates shown in Table 6-5 and Table 6-6,
respectively, we can calculate annual costs by multiplying by estimated sales. Table 6-9
shows projected car sales by manufacturer for model years 2012-2016.  Table 6-10 shows
projected truck sales by manufacturer for model years 2012-2016.  Table 6-11 shows
combined sales by manufacturer for 2012-2016. Table 6-11 shows annual costs attributable
to cars by manufacturer for MYs 2012-2016, Table 6-12 shows the same for trucks, and Table
6-13 shows the same for cars and trucks combined.  Table 6-14 then shows the annual costs
by the entire industry for cars, trucks,  and total for the years 2012 through 2050 with net
present values using both a 3 percent and a 7 percent discount rate.E
E Note that the vehicle compliance costs presented here do not include costs associated with upgrading testing
facilities to accommodate N2O testing. While including those costs would likely have very little impact on the
costs presented here for new vehicle technology, the costs should be included and we intend to do so in the final
rule analysis.
                                          6-7

-------
Regulatory Impact Analysis
                Table 6-9 Estimated Annual Car Sales by Manufacturer (# of Units)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Industry
2012MY
283,471
178,635
1,348,260
149,192
1,484,580
1,155,008
580,538
22,878
304,524
313,489
172,172
64,843
958,696
31,605
86,537
1,697,762
423,433
9,255,624
2013MY
323,191
175,072
1,424,345
135,946
1,620,301
1,352,607
558,975
32,822
309,667
318,669
190,133
61,169
1,031,569
35,813
86,220
1,862,201
458,641
9,977,341
2014MY
352,248
169,046
1,446,097
135,141
1,708,507
1,493,242
562,862
35,534
331,198
338,487
204,335
56,478
1,073,307
38,470
81,480
1,985,033
467,885
10,479,350
201 5MY
371,668
136,583
1,503,175
130,588
1,789,813
1,562,496
590,579
40,586
347,533
340,069
229,562
52,368
1,104,272
36,175
75,965
2,114,273
465,263
10,890,967
2016MY
380,804
138,602
1,511,354
131,022
1,820,234
1,593,092
596,891
41,584
351,081
345,489
235,205
53,459
1,123,486
37,064
77,427
2,154,115
476,699
11,067,608
               Table 6-10 Estimated Annual Truck Sales by Manufacturer (# of Units)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Industry
2012MY
142,861
282,472
1,025,575
88,877
1,209,642
669,948
141,028
42,168
78,789
71,930
109,351
10,098
477,897
20,767
30,391
1,019,375
109,415
5,530,583
2013MY
141,949
244,441
1,107,535
77,593
1,320,438
724,500
135,751
46,256
74,596
71,136
115,684
11,440
478,571
20,639
29,750
1,019,048
100,952
5,720,280
2014MY
144,022
222,545
1,202,141
76,913
1,342,856
687,208
134,274
53,335
72,192
66,919
125,189
9,124
474,558
19,379
30,545
1,045,671
104,023
5,810,895
201 5MY
139,034
112,381
1,230,889
73,425
1,355,820
691,641
132,978
48,583
69,894
59,361
113,366
7,349
454,488
16,822
27,204
1,081,323
102,743
5,717,300
2016MY
135,569
109,674
1,202,442
74,135
1,322,512
675,173
129,763
47,372
68,310
57,998
110,541
7,171
444,471
17,273
26,526
1,057,837
100,186
5,586,953

-------
                           Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-11 Estimated Annual Costs by Manufacturer, including A/C, for Cars (SMillions of 2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Industry
2012MY
$120
$100
$70
$930
$750
$180
$200
$80
$140
$30
$290
$10
$70
$50
$10
$260
$170
$3,460
2013MY
$260
$120
$150
$1,180
$1,040
$390
$260
$120
$180
$40
$460
$30
$90
$60
$20
$530
$350
$5,300
2014MY
$400
$130
$220
$1,370
$1,260
$610
$310
$160
$230
$50
$610
$40
$100
$60
$30
$800
$510
$6,950
201 5MY
$530
$130
$310
$1,690
$1,600
$800
$400
$210
$280
$50
$800
$50
$100
$70
$40
$1,080
$640
$8,850
2016MY
$650
$180
$380
$2,170
$1,760
$960
$440
$260
$330
$60
$1,140
$60
$120
$80
$50
$1,290
$780
$10,710
  Table 6-12 Estimated Annual Costs by Manufacturer, including A/C, for Trucks (SMillions of 2007
                                         dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Industry
2012MY
$60
$140
$40
$580
$440
$130
$80
$30
$10
$0
$170
$0
$60
$20
$10
$120
$30
$1,980
2013MY
$110
$190
$70
$840
$790
$240
$100
$40
$20
$10
$280
$10
$70
$30
$20
$220
$50
$3,090
2014MY
$160
$210
$110
$1,060
$1,040
$330
$120
$50
$30
$10
$360
$10
$80
$30
$20
$310
$70
$3,960
201 5MY
$190
$140
$130
$1,370
$1,440
$410
$150
$70
$40
$10
$470
$10
$90
$30
$30
$410
$80
$5,060
2016MY
$230
$170
$150
$1,790
$2,360
$470
$220
$80
$60
$10
$530
$10
$100
$30
$30
$460
$100
$6,780
                                          6-9

-------
Regulatory Impact Analysis
  Table 6-13 Estimated Annual Costs by Manufacturer, including A/C, for Cars and Trucks Combined
                                   (SMillions of 2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Industry
2012MY
$180
$240
$110
$1,510
$1,190
$310
$280
$110
$150
$30
$460
$10
$130
$70
$20
$380
$200
$5,440
2013MY
$370
$310
$220
$2,020
$1,830
$630
$360
$160
$200
$50
$740
$40
$160
$90
$40
$750
$400
$8,390
2014MY
$560
$340
$330
$2,430
$2,300
$940
$430
$210
$260
$60
$970
$50
$180
$90
$50
$1,110
$580
$10,910
201 5MY
$720
$270
$440
$3,060
$3,040
$1,210
$550
$280
$320
$60
$1,270
$60
$190
$100
$70
$1,490
$720
$13,910
2016MY
$880
$350
$530
$3,960
$4,120
$1,430
$660
$340
$390
$70
$1,670
$70
$220
$110
$80
$1,750
$880
$17,490
        Table 6-14 Annual Sales & Costs for Cars & Trucks (Monetary Values in 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
CAR SALES
9,255,624
9,977,341
10,479,350
10,890,967
11,067,608
11,398,169
11,684,257
11,948,850
12,190,082
12,184,615
12,224,907
12,393,064
12,615,769
12,867,956
13,056,941
13,146,812
13,245,293
13,390,625
13,550,044
13,653,024
13,756,787
13,861,339
13,966,685
TRUCK SALES
5,530,583
5,720,280
5,810,895
5,717,300
5,586,953
5,403,989
5,282,864
5,191,459
5,088,666
5,018,346
4,966,015
4,990,624
5,057,793
5,154,435
5,196,282
5,220,321
5,211,789
5,172,196
5,250,009
5,289,909
5,330,112
5,370,621
5,411,438
CAR COSTS
($MILLIONS)
$3,460
$5,300
$6,950
$8,850
$10,710
$11,030
$11,310
$11,570
$11,800
$11,790
$10,880
$11,030
$11,230
$11,450
$11,620
$11,700
$11,790
$11,920
$12,060
$12,150
$12,250
$12,340
$12,430
TRUCK COSTS
($MILLIONS)
$1,980
$3,090
$3,960
$5,060
$6,780
$6,560
$6,410
$6,300
$6,170
$6,090
$5,540
$5,570
$5,640
$5,750
$5,800
$5,830
$5,820
$5,770
$5,860
$5,900
$5,950
$5,990
$6,040
TOTAL COSTS
($MILLIONS)
$5,440
$8,390
$10,910
$13,910
$17,490
$17,590
$17,720
$17,870
$17,970
$17,880
$16,420
$16,600
$16,870
$17,200
$17,420
$17,530
$17,610
$17,690
$17,920
$18,050
$18,200
$18,330
$18,470
                                            6-10

-------
                          Vehicle Program Costs Including Fuel Consumption Impacts
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
14,072,832
14,179,785
14,287,552
14,396,137
14,505,548
14,615,790
14,726,870
14,838,794
14,951,569
15,065,201
15,179,696
15,295,062
15,411,305
15,528,430
15,646,447
15,765,360


5,452,565
5,494,004
5,535,759
5,577,830
5,620,222
5,662,936
5,705,974
5,749,339
5,793,034
5,837,061
5,881,423
5,926,122
5,971,160
6,016,541
6,062,267
6,108,340


$12,530
$12,620
$12,720
$12,820
$12,910
$13,010
$13,110
$13,210
$13,310
$13,410
$13,510
$13,620
$13,720
$13,820
$13,930
$14,030
$257,680
$141,860
$6,080
$6,130
$6,180
$6,220
$6,270
$6,320
$6,370
$6,420
$6,460
$6,510
$6,560
$6,610
$6,660
$6,710
$6,770
$6,820
$132,320
$74,700
$18,610
$18,750
$18,900
$19,040
$19,180
$19,330
$19,480
$19,630
$19,770
$19,920
$20,070
$20,230
$20,380
$20,530
$20,700
$20,850
$390,000
$216,550
6.2  Cost per Ton of Emissions Reduced

       We have calculated the cost per ton of GHG (CO2 equivalent, or CO2e) reductions
associated with this proposal using the costs shown in Table 6-14 and the emissions
reductions described in Chapter 5. We have calculated the cost per metric ton of GHG
emissions reductions in the years 2020, 2030, 2040, and 2050 using the annual vehicle
compliance costs and emission reductions for each of those years.  The value in 2050 repre-
sents the long-term cost per ton of the emissions reduced.  Note that we have not included the
savings associated with reduced fuel consumption, nor any of the other benefits of this
proposal in the cost per ton calculations.  If we were to include fuel savings in the cost
estimates, the cost per ton would be less than $0, since the fuel savings outweigh the costs
(see Section 6.3 below). With regard to the proposed CH4 and N2O standards, since these
standards would be emissions caps designed to ensure manufacturers do not backslide from
current levels, we have not estimated costs associated with the standards (since the standards
would not require any change from current practices nor do we estimate they would result in
emissions reductions).

       The results for CO2e costs per ton under the proposed vehicle program are shown in
Table 6-15.
                                        6-11

-------
Regulatory Impact Analysis
            Table 6-15 Annual Cost Per Metric Ton of CO2e Reduced, in S2007 dollars
YEAR
2020
2030
2040
2050
COST
($MILLIONS) *
$18,000
$17,900
$19,300
$20,900
CO2-EQUIVLANET
REDUCTION
(MILLION METRIC
TONS)
170
320
420
520
COST PER TON
$110
$60
$50
$40
* Costs here include vehicle compliance costs and do not include any fuel savings (discussed in section 6.3) or
other benefits of this proposal (discussed in Chapter 8).
6.3  Fuel Consumption Impacts

       In this section, we present the impact of the proposed program on fuel consumption
and the consumer savings realized due to the lower fuel consumption.  Chapter 5 provides
more detail on the estimated reduction in the gallons of fuel expected to be consumed as a
result of the proposal.

       The proposed CO2 standards would result in significant improvements in the fuel
efficiency of affected vehicles.  Drivers of those vehicles would see corresponding savings
associated with reduced fuel expenditures. We have estimated the impacts on fuel
consumption for both the proposed tailpipe CO2 standards and the proposed A/C credit
program. To do this, fuel consumption is calculated using both current CO2 emission levels
and the proposed CO2 standards. The difference between these estimates represents the net
savings from the proposed CO2 standards.

       The expected impacts on fuel consumption are shown in Table 6-16. The gallons
shown in the tables reflect impacts from the proposed CO2 standards, including the proposed
A/C credit program, and include increased consumption resulting from the rebound effect.
Using these fuel consumption estimates, we can calculate the  monetized fuel savings
associated with the proposed CO2 standards. To do this, we multiply reduced fuel
consumption in each year by the corresponding estimated average fuel price in that year,
using the reference case taken from the AEO 2009.  AEO is the government consensus
estimate used by NHTSA and many other government agencies to estimate the projected price
of fuel. We have included all fuel taxes in these estimates since these are the prices paid by
consumers. As such, the savings shown reflect savings to the consumer. These results are
also shown in Table 6-16. Note that we present the monetized fuel savings using pre-tax fuel
prices in Chapter 8 of this draft RIA. The fuel savings  based on pre-tax fuel prices reflect the
societal savings in contrast to the consumer savings presented in Table 6-16.  Also in Chapter
8, we present the benefit-cost of the proposal and, for that reason, present the fuel impacts as
negative costs of the program while here we present them as positive savings.
                                        6-12

-------
                       Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-16 Annual Fuel Consumption Impacts of the Proposed Vehicle Standards and A/C Credit
                                   Programs

                          (Monetary values in 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
GALLONS
(MILLIONS)
500
1,300
2,400
3,900
5,900
7,900
9,800
11,600
13,400
15,000
16,600
18,100
19,500
20,900
22,100
23,300
24,300
25,300
26,200
27,000
27,800
28,600
29,400
30,100
30,900
31,600
32,400
33,200
33,900
34,700
35,500
36,300
37,200
38,000
38,900
39,800
40,700
41,600
42,600


FUEL PRICE
INCLUDING
TAXES
($/GALLON)
$2.70
$2.85
$3.00
$3.16
$3.27
$3.39
$3.48
$3.56
$3.62
$3.64
$3.67
$3.69
$3.69
$3.68
$3.72
$3.72
$3.76
$3.87
$3.82
$3.84
$3.86
$3.88
$3.90
$3.92
$3.95
$3.97
$3.99
$4.01
$4.03
$4.05
$4.07
$4.10
$4.12
$4.14
$4.16
$4.19
$4.21
$4.23
$4.25


SAVINGS
($MILLIONS)
$1,400
$3,800
$7,200
$12,400
$19,400
$26,700
$34,100
$41,300
$48,400
$54,600
$60,800
$66,700
$72,000
$76,800
$82,200
$86,500
$91,600
$97,800
$100,000
$103,800
$107,500
$111,100
$114,700
$118,300
$121,900
$125,500
$129,200
$133,000
$136,800
$140,700
$144,700
$148,800
$153,100
$157,400
$161,900
$166,500
$171,200
$176,100
$181,000
$1,850,200
$826,900
                                     6-13

-------
Regulatory Impact Analysis
       As shown in Table 6-16, we are projecting that consumers would realize very large
fuel savings as a result of the standards contained in today's proposal.  There are several ways
to view this value.  Some, as demonstrated below in Chapter 8 of this draft RIA, view these
fuel savings as a reduction in the cost of owning a vehicle, whose full benefits consumers
realize. This approach assumes that, regardless how consumers in fact make their decisions
on how much fuel economy to purchase, they will gain these fuel savings. Another view says
that consumers do not necessarily value fuel savings as equal to the results of this calculation.
Instead, consumers may either undervalue or overvalue fuel economy relative to these
savings, based on their personal preferences.  This issue is discussed further in Section 8.1.2
of this draft RIA.

       If we limit the analysis to the five model years 2012-2016—in other words, the fuel
consumption savings during the lifetimes of those five model years, the results would be as
shown in Table 6-17.
 Table 6-17 Annual Fuel Savings for 2012-2016 MY Vehicles Using Pre-tax Fuel Prices (SMillions of 2007
                                        dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2012MY
$1,300
$1,300
$1,400
$1,400
$1,400
$1,400
$1,400
$1,300
$1,200
$1,100
$1,100
$900
$800
$600
$500
$400
$300
$300
$200
$200
$100
$100
$100
$100
$100
$100
$0
$0
$0
$0
$0
$0
2013MY

$2,100
$2,200
$2,200
$2,200
$2,200
$2,200
$2,100
$2,000
$1,900
$1,700
$1,600
$1,400
$1,200
$1,000
$800
$700
$600
$400
$300
$300
$200
$200
$200
$100
$100
$100
$100
$100
$0
$0
$0
2014MY


$3,100
$3,200
$3,200
$3,200
$3,200
$3,100
$3,000
$2,800
$2,600
$2,400
$2,200
$2,000
$1,700
$1,400
$1,100
$1,000
$800
$600
$500
$400
$300
$300
$200
$200
$200
$100
$100
$100
$100
$100
2015MY



$4,600
$4,700
$4,700
$4,700
$4,600
$4,400
$4,200
$4,000
$3,700
$3,400
$3,100
$2,800
$2,300
$2,000
$1,700
$1,300
$1,100
$900
$700
$600
$500
$400
$300
$300
$200
$200
$100
$100
$100
2016MY




$6,500
$6,600
$6,600
$6,500
$6,300
$6,000
$5,700
$5,400
$5,000
$4,600
$4,200
$3,700
$3,200
$2,700
$2,200
$1,800
$1,500
$1,200
$1,000
$800
$600
$500
$400
$400
$300
$300
$200
$200
SUM
$1,300
$3,500
$6,700
$11,500
$18,100
$18,100
$18,000
$17,600
$17,000
$16,100
$15,100
$14,100
$12,800
$11,500
$10,200
$8,700
$7,300
$6,200
$5,000
$4,000
$3,300
$2,600
$2,100
$1,700
$1,400
$1,200
$1,000
$800
$700
$600
$500
$400
                                         6-14

-------
                          Vehicle Program Costs Including Fuel Consumption Impacts
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
$0
$0
$0
$0
$0
$0
$0
$15,600
$12,100
$0
$0
$0
$0
$0
$0
$0
$24,400
$19,000
$100
$100
$0
$0
$0
$0
$0
$34,800
$27,200
$100
$100
$100
$100
$100
$0
$0
$49,800
$38,900
$100
$100
$100
$100
$100
$100
$100
$68,500
$53,700
$300
$300
$300
$200
$200
$200
$100
$193,100
$150,900
6.4  Vehicle Program Cost Summary

       The vehicle program costs consist of the vehicle compliance costs and the fuel savings
(fuel savings are expressed here as negative fuel costs) that would result from the reduction in
fuel consumption. These costs are summarized in Table 6-18.

 Table 6-18 Annual Vehicle Program Costs Including Fuel Costs Using Post-Tax Fuel Prices (SMillions of
                                     2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
VEHICLE
COMPLIANCE
COSTS
$5,400
$8,400
$10,900
$13,900
$17,500
$17,600
$17,700
$17,900
$18,000
$17,900
$16,400
$16,600
$16,900
$17,200
$17,400
$17,500
$17,600
$17,700
$17,900
$18,100
$18,200
$18,300
$18,500
$18,600
$18,800
$18,900
$19,000
$19,200
FUEL COSTS
(NEGATIVE COSTS
ARE SAVINGS)
-$1,400
-$3,800
-$7,200
-$12,400
-$19,400
-$26,700
-$34,100
-$41,300
-$48,400
-$54,600
-$60,800
-$66,700
-$72,000
-$76,800
-$82,200
-$86,500
-$91,600
-$97,800
-$100,000
-$103,800
-$107,500
-$111,100
-$114,700
-$118,300
-$121,900
-$125,500
-$129,200
-$133,000
TOTAL COSTS
(NEGATIVE COSTS
ARE SAVINGS)
$4,000
$4,600
$3,700
$1,500
-$1,900
-$9,100
-$16,400
-$23,400
-$30,400
-$36,700
-$44;400
-$50,100
-$55,100
-$59,600
-$64,800
-$69,000
-$74,000
-$80,100
-$82,100
-$85,700
-$89,300
-$92,800
-$96,200
-$99,700
-$103,100
-$106,600
-$110,200
-$113,800
                                        6-15

-------
Regulatory Impact Analysis
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
$19,300
$19,500
$19,600
$19,800
$19,900
$20,100
$20,200
$20,400
$20,500
$20,700
$20,900
$390,000
$216,600
-$136,800
-$140,700
-$144,700
-$148,800
-$153,100
-$157,400
-$161,900
-$166,500
-$171,200
-$176,100
-$181,000
-$1,850,200
-$826,900
-$117,500
-$121,200
-$125,100
-$129,000
-$133,200
-$137,300
-$141,700
-$146,100
-$150,700
-$155,400
-$160,100
-$1,460,200
-$610,300
                                      6-16

-------
                          Vehicle Program Costs Including Fuel Consumption Impacts









References



All references can be found in the EPA DOCKET: EPA-HQ-OAR-2009-0472.
1 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. August 2009.




2 US EPA 2009. Cost effective achieved levels spreadsheet.
                                        6-17

-------

-------
                                                  Environmental and Health Impacts

CHAPTER 7: Environmental and Health Impacts

7.1  Health and Environmental Effects of Non-GHG Pollutants

7.1.1 Health Effects Associated with Exposure to Pollutants

       In this section we will discuss the health effects associated with non-GHG pollutants,
specifically: paniculate matter, ozone, nitrogen oxides (NOx), sulfur oxides (SOX), carbon
monoxide and air toxics. These pollutants would not be directly regulated by the proposed
standards, but the proposed standards would affect emissions of these  pollutants and
precursors.  Reductions in these pollutants would be co-benefits of this proposal (that is,
benefits in addition to the benefits of reduced GHGs).

7.1.1.1   Participate Matter

7.1.1.1.1  Background

       Paniculate matter (PM) is a generic term for a broad class of chemically and
physically diverse substances. It can be principally characterized as discrete particles that
exist in the condensed (liquid or solid) phase spanning several orders of magnitude in size.
Since 1987, EPA has delineated that subset of inhalable particles small enough to penetrate to
the thoracic region (including the tracheobronchial and alveolar regions) of the respiratory
tract (referred to as thoracic particles). Current national ambient air quality standards
(NAAQS) use PM2 5 as the indicator for fine particles (with PM2 5 referring to particles with a
nominal mean aerodynamic diameter less than or equal to 2.5 um), and use PMio as the
indicator for purposes of regulating the coarse fraction of PMio (referred to as thoracic  coarse
particles or coarse-fraction particles; generally including particles with a nominal mean
aerodynamic diameter greater than 2.5 um and less than or equal to 10 um, or PMi0-2.5).
Ultrafine particles are a subset of fine particles, generally less than 100 nanometers (0.1 um)
in aerodynamic diameter.

       Particles span many sizes and shapes and consist of hundreds of different chemicals.
Particles originate from sources and are also formed through atmospheric chemical reactions;
the former are often referred to as "primary" particles, and the latter as "secondary" particles.
In addition, there are also physical, non-chemical reaction mechanisms that contribute to
secondary particles.  Particle pollution also varies by time of year and location and is affected
by several weather-related factors, such as temperature, clouds, humidity, and wind.  A
further layer of complexity comes from a particle's ability to shift between solid/liquid and
gaseous phases, which is influenced by concentration, meteorology, and temperature.

       Fine particles are produced primarily by combustion processes and by transformations
of gaseous emissions (e.g., SOx, NOx and VOCs) in the atmosphere. The chemical and
physical properties of PM25 may vary greatly with time, region, meteorology and source
category. Thus, PM2 5 may include a complex mixture of different pollutants including
sulfates, nitrates, organic compounds, elemental carbon and metal compounds. These
particles can remain in the atmosphere for days to weeks and travel through the atmosphere
hundreds to thousands of kilometers.1
                                         7-1

-------
Draft Regulatory Impact Analysis
7.1.1.1.2  Health Effects ofPM

        This section provides a summary of the health effects associated with exposure to
ambient concentrations of PM.A The information in this section is based on the data and
conclusions in the PM Air Quality Criteria Document (PM AQCD) and PM Staff Paper
prepared by the U.S. Environmental Protection Agency  (EPA).B'2'3 We also present
additional recent studies published after the cut-off date for the PM AQCD.4'0  Taken together
this information supports the conclusion that exposure to ambient concentrations of PM are
associated with adverse health effects.

         7.1.1.1.2.1  Short-term Exposure Mortality and Morbidity Studies

        As discussed in the PM AQCD, short-term exposure to PM2 5 is associated with
premature mortality from cardiopulmonary  diseases,5 hospitalization and emergency
department visits for cardiopulmonary diseases,6 increased respiratory symptoms,7 decreased
lung function8 and physiological changes or biomarkers for cardiac changes.9'10 In addition,
the PM AQCD described a limited body  of new evidence from epidemiologic studies for
potential relationships between  short term exposure to PM and health endpoints such as low
birth weight, preterm birth, and neonatal and infant mortality.11

       Among the studies of effects associated with short-term exposure to PM2 5, several
specifically address the contribution of mobile sources to short-term PM2 s-related effects on
premature mortality. The results from these studies generally indicated that several
combustion-related fine particle source-types are likely associated with mortality,  including
A Personal exposure includes contributions from many different types of particles, from many sources, and in
many different environments. Total personal exposure to PM includes both ambient and nonambient
components; and both components may contribute to adverse health effects.
B The PM NAAQS is currently under review and the EPA is considering all available science on PM health
effects, including information which has been published since 2004, in the development of the upcoming PM
Integrated Science Assessment Document (ISA).  A second draft of the PM ISA was completed in July 2009 and
was submitted for review by the Clean Air Scientific Advisory Committee (CAS AC) of EPA's Science Advisory
Board.  Comments from the general public have also been requested. For more information, see
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=210586.
c These additional studies are included in the 2006 Provisional Assessment of Recent Studies on Health Effects
of Particulate Matter Exposure. The provisional assessment did not and could not (given a very short timeframe)
undergo the extensive critical review by CASAC and the public, as did the PM AQCD. The provisional
assessment found that the "new" studies expand the scientific information and provide important insights on the
relationship between PM exposure and health effects of PM. The provisional assessment also found that "new"
studies generally strengthen the evidence  that acute and chronic exposure to fine particles and acute exposure to
thoracic coarse particles are associated with health effects. Further, the provisional science assessment found
that the results reported in the studies did not dramatically diverge from previous findings, and taken in context
with the findings of the AQCD, the new information and findings did not materially change any of the broad
scientific conclusions  regarding the health effects of PM exposure made in the AQCD. However, it is important
to note that this assessment was limited to screening, surveying, and preparing a provisional assessment of these
studies. For reasons outlined in Section I.C of the preamble for the final PM NAAQS rulemaking in 2006 (see
71 FR 61148-49, October 17, 2006), EPA based its NAAQS decision on the science presented in the 2004
AQCD.
                                             7-2

-------
                                                   Environmental and Health Impacts

motor vehicle emissions as well as other sources.12 The analyses incorporate source
apportionment tools into short-term exposure studies and are briefly mentioned here.
Analyses incorporating source apportionment by factor analysis with daily time-series studies
of daily death rates indicated a relationship between mobile source PM2 5 and
mortality.13'14'15'16 Another recent study in 14 U.S. cities examined the effect of PMi0
exposures on daily hospital admissions for cardiovascular disease. This study found that the
effect of PMio was significantly greater in areas with a larger proportion of PMi0 coming from
motor vehicles, indicating that PMio from these sources may have a greater effect on the
toxicity of ambient PMio when compared with other sources.17 These studies provide
evidence that PM-related emissions, specifically from mobile sources, are associated with
adverse health effects.

         7.1.1.1.2.2   Long-term Exposure Mortality and Morbidity Studies

        Long-term exposure to ambient PM2 5 is associated with premature mortality from
cardiopulmonary diseases and lung cancer,18 and effects on the respiratory system such as
decreased lung function or the development of chronic respiratory disease.19 Of specific
importance, the PM AQCD  also noted that the PM components of gasoline and diesel engine
exhaust represent one class of hypothesized likely important contributors to the observed
ambient PM-related increases in lung cancer incidence and mortality.20

       The PM AQCD and PM Staff Paper emphasized the results of two long-term
epidemiologic studies, the Six Cities and American Cancer Society (ACS) prospective cohort
studies, based on several factors - the large air quality data set for PM in the Six Cities Study,
the fact that the study populations were  similar to the general population, and the fact that
these studies have undergone extensive reanalysis.21'22'23'2425'26  These studies indicate that
there are positive associations for all-cause, cardiopulmonary, and lung cancer mortality with
long-term exposure to PM25. One analysis of a subset of the  ACS cohort data, which was
published after the PM AQCD was finalized but in time for the 2006 Provisional Assessment,
found a larger association than had previously been reported between long-term PM2 5
exposure and mortality  in the Los Angeles area using a new exposure estimation method that
accounted for variations in concentration within the city.27

       As discussed in the PM AQCD, the morbidity studies that combine the features of
cross-sectional and cohort studies provide the best evidence for chronic exposure effects.
Long-term studies evaluating the effect of ambient PM on children's development have
shown some evidence indicating effects of PM25 and/or PMio on reduced lung function
growth.28 In another recent publication included in the 2006 Provisional Assessment,
investigators in southern California reported the results of a cross-sectional study of outdoor
PM2 5 and a measure of atherosclerosis development in the Los Angeles basin.29 The study
found significant associations between ambient residential PM2 5 and carotid intima-media
thickness (CIMT), an indicator of subclinical atherosclerosis, an underlying factor in
cardiovascular disease.

7.1.1.2  Ozone

7.1.1.2.1   Background

                                          7-3

-------
Draft Regulatory Impact Analysis

       Ground-level ozone pollution is typically formed by the reaction of VOCs and NOx
in the lower atmosphere in the presence of heat and sunlight. These pollutants, often referred
to as ozone precursors, are emitted by many types of pollution sources such as highway and
nonroad motor vehicles and engines, power plants, chemical plants, refineries, makers of
consumer and commercial products, industrial facilities, and smaller area sources.

       The science of ozone formation, transport, and accumulation is complex.  Ground-
level ozone is produced and destroyed in a cyclical set of chemical reactions, many of which
are sensitive to temperature and sunlight.  When ambient temperatures and sunlight levels
remain high for several days  and the air is relatively stagnant, ozone and its precursors can
build up and result in more ozone than typically occurs on a single high-temperature day.
Ozone can be transported hundreds of miles downwind of precursor emissions, resulting in
elevated ozone levels even in areas with low VOC or NOx emissions.

       The highest levels of ozone are produced when both VOC and NOx emissions are
present in significant quantities on clear summer days.  Relatively small amounts of NOx
enable ozone to form rapidly when VOC levels are relatively high, but ozone production is
quickly limited by removal of the NOx.  Under these conditions NOx reductions are highly
effective in reducing ozone while VOC reductions have little effect.  Such conditions are
called "NOx-limited." Because the contribution of VOC emissions from biogenic (natural)
sources to local ambient ozone concentrations can be significant, even some areas where man-
made VOC emissions are relatively low can be NOx-limited.

       Ozone concentrations in an area also can be lowered by the reaction of nitric oxide
(NO) with ozone, forming nitrogen dioxide (NO2); as the air moves downwind and the cycle
continues, the NO2 forms additional ozone. The importance of this reaction depends, in part,
on the relative concentrations of NOx, VOC, and ozone, all of which change with time and
location. When NOx levels are relatively high and VOC levels relatively low, NOx forms
inorganic nitrates (i.e., particles) but relatively little ozone. Such conditions are called "VOC-
limited". Under these conditions, VOC reductions are effective in reducing ozone, but NOx
reductions can actually increase local ozone under certain circumstances. Even in VOC-
limited urban areas,  NOx reductions are not expected to increase ozone levels if the NOx
reductions are sufficiently large.

       Rural areas are usually NOx-limited, due to the relatively large amounts of biogenic
VOC emissions in such areas. Urban areas can be either VOC- or NOx-limited, or a mixture
of both, in which ozone levels exhibit moderate sensitivity to changes in either pollutant.

7.1.1.2.2   Health Effects of Ozone

       Exposure to ambient ozone contributes to a wide range of adverse health effects.0
These health  effects are well documented and are critically assessed in the EPA ozone air
quality criteria document (ozone AQCD) and EPA staff paper.30'31 We are relying on the data
D Human exposure to ozone varies over time due to changes in ambient ozone concentration and because people
move between locations which have notable different ozone concentrations. Also, the amount of ozone
delivered to the lung is not only influenced by the ambient concentrations but also by the individuals breathing
route and rate.

                                         7-4

-------
                                                    Environmental and Health Impacts

and conclusions in the ozone AQCD and staff paper, regarding the health effects associated
with ozone exposure.

       Ozone-related health effects include lung function decrements, respiratory symptoms,
aggravation of asthma, increased hospital and emergency room visits, increased asthma
medication usage, and a variety of other respiratory effects. Cellular-level effects, such as
inflammation of lungs, have been documented as well. In addition, there is suggestive
evidence of a contribution of ozone to cardiovascular-related morbidity and highly suggestive
evidence that short-term ozone exposure directly or indirectly contributes to non-accidental
and cardiopulmonary-related mortality, but additional research is needed to clarify the
underlying mechanisms causing these effects. In a recent report on the estimation of ozone-
related premature mortality published by the National Research Council (NRC), a panel of
experts and reviewers concluded that short-term exposure to ambient ozone is likely to
contribute to premature deaths and that ozone-related mortality should be included in
estimates of the health benefits of reducing ozone exposure.32 People who are more
susceptible to effects associated with exposure to ozone can include children, asthmatics and
the elderly. Those with greater exposures to ozone, for instance due to time spent outdoors
(e.g., children and outdoor workers), are also of concern.

       Based on a large number of scientific studies, EPA has identified several key health
effects associated with exposure to levels of ozone found today in many areas of the country.
Short-term (1 to 3 hours) and prolonged exposures (6 to 8 hours) to ambient ozone
concentrations have been linked to lung function decrements,  respiratory symptoms, increased
hospital admissions and emergency room visits for respiratory problems.33'34'35'36> 37'38
Repeated exposure to ozone can increase susceptibility to respiratory infection and lung
inflammation and can aggravate preexisting respiratory diseases, such as asthma.39'40> 41> 42> 43
Repeated exposure to sufficient concentrations of ozone can also cause inflammation of the
lung, impairment of lung defense mechanisms, and possibly irreversible changes in lung
structure, which over time could affect premature aging of the lungs and/or the development
of chronic respiratory illnesses, such as emphysema and chronic bronchitis.44'45> 46'47

       Children and adults who are outdoors and active during the summer months, such as
construction workers, are among those most at risk of elevated ozone exposures.48 Children
and outdoor workers tend to have higher ozone exposure because they typically are active
outside, working, playing and exercising, during times of day  and seasons (e.g., the summer)
when ozone levels are highest.49  For example, summer camp  studies in the Eastern United
States and Southeastern Canada have reported statistically significant reductions in lung
function in children who are active outdoors.50'51'52'53'54> 55'56'57  Further, children are more at
risk of experiencing health effects from ozone exposure than adults because their respiratory
systems are still developing. These individuals (as well as people with respiratory illnesses,
such as asthma, especially asthmatic children) can experience reduced lung function and
increased respiratory symptoms, such as chest pain and cough, when exposed to relatively low
ozone levels during prolonged periods of moderate exertion.58'59> 60> 61
                                          7-5

-------
Draft Regulatory Impact Analysis



7.1.1.3  Nitrogen Oxides and Sulfur Oxides

7.1.1.3.1  Background

        Sulfur dioxide (SCh), a member of the sulfur oxide (SOX) family of gases, is formed
from burning fuels containing sulfur (e.g., coal or oil derived), extracting gasoline from oil, or
extracting metals from ore. Nitrogen dioxide (NCh) is a member of the nitrogen oxide (NOx)
family of gases. Most NCh is formed in the air through the oxidation of nitric oxide (NO)
emitted when fuel is burned at a high temperature.

        862 andNCh can dissolve in water vapor and further oxidize to form sulfuric and
nitric acid which react with ammonia to form sulfates and nitrates, both of which are
important components of ambient PM. The health effects of ambient PM are discussed in
Section 7.1.1.1.2.  NOx along with non-methane hydrocarbons (NMHC) are the two major
precursors of ozone.  The health effects of ozone  are covered in Section 7.1.1.2.2.

7.1.1.3.2  Health Effects of Sulfur Oxides

       Information on the health effects of SO2 can be found in the U.S. Environmental
Protection Agency Integrated Science Assessment for Sulfur Oxides.62 SO2 has long been
known to cause adverse respiratory  health effects, particularly among individuals with
asthma. Other potentially sensitive  groups include children and the elderly. During periods of
elevated ventilation, asthmatics may experience symptomatic bronchoconstriction within
minutes of exposure. Following an extensive evaluation of health evidence from
epidemiologic  and laboratory studies, the EPA has concluded that there is a causal
relationship between respiratory health effects and short-term exposure to SO2.  Separately,
based on an evaluation of the epidemiologic evidence of associations between short-term
exposure to SO2 and mortality, the EPA has concluded that the overall evidence is suggestive
of a causal relationship between short-term exposure to SO2 and mortality.

7.1.1.3.3  Health Effects of Nitrogen Oxides

       Information on the health effects of NO2 can be found in the U.S. Environmental
Protection Agency Integrated Science Assessment (ISA) for Nitrogen Oxides.63  The U.S.
EPA has concluded that the findings of epidemiologic, controlled human exposure, and
animal toxicological studies provide evidence that is sufficient to infer a likely causal
relationship between respiratory effects and short-term NO2 exposure. The ISA concludes
that the  strongest evidence for such a relationship comes from epidemiologic studies of
respiratory effects including symptoms, emergency  department visits, and hospital
admissions.  The ISA also draws two broad conclusions regarding airway responsiveness
following NO2 exposure.  First, the  ISA concludes that NO2 exposure may enhance the
sensitivity to allergen-induced decrements in lung function and increase the allergen-induced
airway inflammatory response at exposures as low as 0.26 ppm NO2 for 30 minutes.  Second,
exposure to NO2 has been found to enhance the inherent responsiveness of the airway to
subsequent nonspecific challenges in controlled human exposure studies of asthmatic
                                         7-6

-------
                                                    Environmental and Health Impacts

subjects.  Enhanced airway responsiveness could have important clinical implications for
asthmatics since transient increases in airway responsiveness following NCh exposure have
the potential to increase symptoms and worsen asthma control. Together, the epidemiologic
and experimental data sets form a plausible, consistent, and coherent description of a
relationship between NO2 exposures and an array of adverse health effects that range from the
onset of respiratory symptoms to hospital admission.

       Although the weight of evidence supporting a causal relationship is somewhat less
certain than that associated with respiratory morbidity, NO2 has also been linked to other
health endpoints.  These include all-cause (nonaccidental) mortality, hospital admissions or
emergency department visits for cardiovascular disease, and decrements in lung function
growth associated with chronic exposure.

7.1.1.4   Carbon Monoxide

        We are relying on the data and conclusions in the EPA Air Quality Criteria Document
for CO (CO Criteria Document), which was published in 2000, regarding the health effects
associated with CO exposure.E'64 Carbon monoxide enters the bloodstream through the lungs
and forms carboxyhemoglobin (COHb), a compound that inhibits the blood's capacity to
carry oxygen to organs and tissues.65'66  Carbon monoxide has long been known to have
substantial adverse effects on human health, including toxic effects on blood and tissues, and
effects on organ functions. Although there are effective compensatory increases in blood
flow to the brain, at some concentrations of COHb somewhere above 20 percent, these
compensations fail to maintain sufficient oxygen delivery, and metabolism declines.67  The
subsequent hypoxia in brain tissue then produces behavioral effects, including decrements in
continuous performance and reaction time.68

       Carbon monoxide has been linked to increased risk for people with heart disease,
reduced visual perception, cognitive functions and aerobic capacity, and possible fetal
effects.69  Persons with heart disease are especially sensitive to CO poisoning and may
experience chest pain if they breathe the gas while exercising.70 Infants, elderly persons, and
individuals with respiratory diseases are also particularly sensitive. Carbon monoxide can
affect healthy individuals, impairing exercise capacity, visual perception, manual dexterity,
learning functions, and ability to  perform complex tasks.71

       Several epidemiological studies have shown a link between CO  and  premature
morbidity (including angina, congestive heart failure, and other cardiovascular diseases).
Several studies in the United States and Canada have also reported an association between
ambient CO exposures and frequency of cardiovascular hospital admissions, especially for
congestive heart failure (CHF).  An association between ambient CO exposure and mortality
has also been reported in epidemiological studies, though not as consistently or specifically as
with CHF admissions.  EPA reviewed these studies as part of the CO Criteria Document
E The CO NAAQS is currently under review and the EPA is considering all available science on CO health
effects, including information which has been published since 2000, in the development of the upcoming CO
Integrated Science Assessment Document (ISA). A first draft of the CO ISA was completed in March 2009 and
was submitted for review by the Clean Air Scientific Advisory Committee (CASAC) of EPA's Science Advisory
Board. For more information, see http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=203935.

                                          7-7

-------
Draft Regulatory Impact Analysis

review process and noted the possibility that the average ambient CO levels used as exposure
indices in the epidemiology studies may be surrogates for ambient air mixes impacted by
combustion sources and/or other constituent toxic components of such mixes. More research
will be needed to better clarify CD's role.72

7.1.1.5   Air Toxics

        Motor vehicle emissions contribute to ambient levels of air toxics known or suspected
as human or animal carcinogens, or that have noncancer health effects.  The population
experiences an elevated risk of cancer and other noncancer health effects from exposure to air
toxics.73 These compounds include, but are not limited to, benzene, 1,3-butadiene,
formaldehyde, acetaldehyde, acrolein, polycyclic organic matter (POM), and naphthalene.
These compounds,  except acetaldehyde, were identified as national or regional risk drivers in
the 2002 National-scale Air Toxics Assessment (NATA) and have significant inventory
contributions from  mobile sources.
      Table 7-1 Mobile Source Inventory Contribution to 2002 Emissions of NATA Risk Drivers
2002 NATA Risk Driver
Benzene
1,3 -Butadiene
Formaldehyde
Acrolein
Polycyclic organic matter (POM)*
Naphthalene
Diesel PM and Diesel exhaust
organic gases
Percent of National
Emissions Attributable
to All Mobile Sources
59%
58%
43%
18%
6%
35%
100%
Percent of National
Emissions Attributable
to Light-Duty Vehicles
41%
37%
19%
9%
3%
22%
1%
   " This table is generated from data contained in the pollutant specific Microsoft Access database files found
   in the State-Specific Emission by County section of the 2002 NATA webpage
   (http://www.epa.gov/ttn/atw/nata2002/tables.html) and data from the 2002 National Emissions Inventory
   (NEI; http://www.epa.gov/ttn/chief/net/2002inventory.html), which is the underlying basis for the emissions
   used in the 2002 NATA (http://www.epa.gov/ttn/atw/nata2002/methods.html).
   6 This POM inventory includes the 15 POM compounds: benzo[b]fluoranthene, benz[a]anthracene, indeno(l,2,3-
   c,d)pyrene, benzo[k]fluoranthene, chrysene, benzo[a]pyrene, dibenz(a,h)anthracene, anthracene, pyrene,
   benzo(g,h,i)perylene, fluoranthene, acenaphthylene, phenanthrene, fluorine, and acenaphthene.

       According to NATA for 2002, mobile sources were responsible for 47 percent of
outdoor toxic emissions, over 50 percent of the cancer risk, and over 80 percent of the
noncancer hazard.  Benzene is the largest contributor to cancer risk of all 124 pollutants
quantitatively assessed in the 2002 NATA and mobile sources were responsible for 59 percent
of benzene emissions in 2002.  In 2007, EPA finalized vehicle and fuel controls that address
this public health risk; it will reduce total emissions of mobile source air toxics by 330,000
tons in 2030, including 61,000 tons of benzene.74

-------
                                                    Environmental and Health Impacts

       Noncancer health effects can result from chronic/ subchronic,G or acute11 inhalation
exposures to air toxics, and include neurological, cardiovascular, liver, kidney, and respiratory
effects as well as effects on the immune and reproductive systems. According to the 2002
NAT A, nearly the  entire U.S. population was exposed to an average concentration of air
toxics that has the potential for adverse noncancer respiratory health effects. This will
continue to be the case in 2030, even though toxics concentrations will be lower. Mobile
sources were responsible for over 80 percent of the noncancer (respiratory) risk from outdoor
air toxics in 2002.  The majority of this risk was from exposure to acrolein. The confidence in
the RfC for acrolein is medium and confidence in NATA estimates of population noncancer
hazard from ambient exposure to this pollutant is low.75'76

       The NATA modeling framework has a number of limitations which prevent its use as
the sole basis for setting regulatory standards. These limitations and uncertainties are
discussed on the 2002 NATA website.77 Even so, this modeling framework is very useful in
identifying air toxic pollutants and sources of greatest concern, setting regulatory priorities,
and informing the decision making process.

7.1.1.5.1   Benzene

        The EPA's IRIS database lists benzene as a known human carcinogen (causing
leukemia) by all routes of exposure, and concludes that exposure is associated with additional
health effects, including genetic changes in both humans and animals and increased
proliferation of bone marrow cells in mice.78'79'80 EPA states in its IRIS database that data
indicate a causal relationship between benzene exposure and acute lymphocytic leukemia and
suggest a relationship between benzene exposure and chronic non-lymphocytic  leukemia and
chronic lymphocytic leukemia.  The International Agency for Research on Carcinogens
(IARC) has determined that benzene is a human carcinogen and the U.S. Department of
Health and Human Services (DHHS) has characterized benzene as a known human
           01 QO
carcinogen. '

       A number of adverse noncancer health effects including blood disorders, such as
preleukemia and aplastic anemia,  have also been associated with long-term exposure to
benzene.83'84  The most sensitive noncancer effect observed in humans, based on current data,
is the depression of the absolute lymphocyte count in blood.85'86 In addition, recent work,
including studies sponsored by the Health Effects Institute (HEI), provides evidence that
biochemical responses are occurring at lower levels of benzene exposure than previously
known.87'88'89'90 EPA's IRIS program has not yet evaluated these new data.

7.1.1.5.2  1,3-Butadiene
F Chronic exposure is defined in the glossary of the Integrated Risk Information (IRIS) database
(http://www.epa.gov/iris) as repeated exposure by the oral, dermal, or inhalation route for more than
approximately 10% of the life span in humans (more than approximately 90 days to 2 years in typically used
laboratory animal species).
0 Defined in the IRIS database as exposure to a substance spanning approximately 10% of the lifetime of an
organism.
H Defined in the IRIS database as exposure by the oral, dermal, or inhalation route for 24 hours or less.

                                          7-9

-------
Draft Regulatory Impact Analysis

       EPA has characterized 1,3-butadiene as carcinogenic to humans by inhalation.91'92
The IARC has determined that 1,3-butadiene is a human carcinogen and the U.S. DHHS has
characterized 1,3-butadiene as a known human carcinogen.93'94 There are numerous studies
consistently demonstrating that 1,3-butadiene is metabolized into genotoxic metabolites by
experimental animals and humans.  The specific mechanisms of 1,3-butadiene-induced
carcinogenesis are unknown; however, the scientific evidence strongly suggests that the
carcinogenic effects are mediated by genotoxic metabolites. Animal data suggest that females
may be more sensitive than males for cancer effects associated with 1,3-butadiene exposure;
there are insufficient data in humans from which to draw conclusions about sensitive
subpopulations.   1,3-butadiene also causes a variety of reproductive and developmental
effects in mice; no human data on these effects are available. The most sensitive effect was
ovarian atrophy  observed in a lifetime bioassay of female mice.95

7.1.1.5.3   Formaldehyde

       Since 1987, EPA has classified formaldehyde as a probable human carcinogen based
on evidence in humans and in rats, mice, hamsters, and monkeys.96 EPA is currently
reviewing recently published epidemiological data. For instance, research conducted by the
National Cancer Institute (NCI) found an increased risk of nasopharyngeal cancer and
lymphohematopoietic malignancies such as leukemia among workers exposed to
formaldehyde.97'98  In an analysis of the lymphohematopoietic cancer mortality from an
extended follow-up of these workers, NCI confirmed an association between
lymphohematopoietic cancer risk and peak exposures.99 A recent National Institute of
Occupational Safety and Health (NIOSH) study of garment workers also found increased risk
of death due to leukemia among workers exposed to formaldehyde.100 Extended follow-up of
a cohort of British chemical workers did not find evidence of an increase in nasopharyngeal or
lymphohematopoietic cancers, but a continuing statistically significant excess in lung cancers
was reported.10

      In the past 15 years there has been substantial research on the inhalation dosimetry for
formaldehyde in rodents and primates by the CUT Centers for Health Research (formerly the
Chemical Industry Institute of Toxicology), with a focus on use of rodent data for refinement
of the quantitative cancer dose-response assessment.102'103'104 CIIT's risk assessment of
formaldehyde incorporated mechanistic and dosimetric information on formaldehyde.
However, it should be noted that recent research published by EPA indicates that when two-
stage modeling assumptions are varied, resulting dose-response estimates can vary by several
orders of magnitude.105'106'107'108  These findings are not supportive of interpreting the CUT
model results as providing a conservative (health protective) estimate of human risk. EPA
research also examined the contribution of the two-stage modeling for formaldehyde towards
characterizing the relative weights of key events in the mode-of-action of a carcinogen. For
example, the model-based inference in the published CUT study that formaldehyde's direct
mutagenic action is not relevant to the compound's tumorigenicity was found not to hold
under variations of modeling assumptions.

      Based on the developments of the last decade, in 2004, the working group of the IARC
concluded that formaldehyde is carcinogenic to humans (Group 1), on the basis of sufficient
evidence in humans and sufficient evidence in experimental animals - a higher classification

                                        7-10

-------
                                                   Environmental and Health Impacts

than previous I ARC evaluations. After reviewing the currently available epidemiological
evidence, the IARC (2006) characterized the human evidence for formaldehyde
carcinogenicity as "sufficient," based upon the data on nasopharyngeal cancers; the
epidemiologic evidence on leukemia was characterized as "strong."109 EPA is reviewing the
recent work cited above from the NCI and NIOSH, as well as the analysis by the CUT Centers
for Health Research and other studies, as part of a reassessment of the human hazard and
dose-response associated with formaldehyde.

       Formaldehyde exposure also causes a range of noncancer health effects, including
irritation of the eyes (burning and watering of the eyes), nose and throat. Effects from
repeated exposure in humans include respiratory tract irritation, chronic bronchitis and nasal
epithelial lesions such as metaplasia and loss of cilia. Animal studies suggest that
formaldehyde may also cause airway inflammation - including eosinophil infiltration into the
airways. There are several studies that suggest that formaldehyde may increase the risk of
asthma - particularly in the young.110'111

7.1.1.5.4   Acetaldehyde

        Acetaldehyde is classified in EPA's IRIS database as a probable human carcinogen,
based on nasal tumors in rats, and is considered toxic by the inhalation, oral, and intravenous
routes.112 Acetaldehyde is reasonably anticipated to be a human carcinogen by the U.S.
DHHS in the 11th Report on Carcinogens and is classified as possibly carcinogenic to humans
(Group 2B) by the IARC.113'114 EPA is currently conducting a reassessment of cancer risk
from inhalation exposure to acetaldehyde.

       The primary noncancer effects of exposure to acetaldehyde vapors  include irritation of
the eyes, skin, and respiratory tract.115  In short-term (4 week) rat studies, degeneration of
olfactory epithelium was observed at various concentration levels of acetaldehyde
exposure.116'117 Data from these studies were used by EPA to develop an inhalation reference
concentration. Some asthmatics have been shown to be a sensitive subpopulation to
decrements in functional expiratory volume (FEV1 test) and bronchoconstriction upon
acetaldehyde inhalation.118 The agency is currently conducting a reassessment of the health
hazards from inhalation exposure to acetaldehyde.

7.1.1.5.5   Acrolein

        EPA determined in 2003 that the human carcinogenic potential of acrolein could not
be determined because the available data were inadequate. No information was available on
the carcinogenic effects of acrolein in humans and the animal data provided inadequate
evidence of carcinogenicity.119 The IARC determined in 1995 that acrolein was not
classifiable as to its carcinogenicity in humans.120

       Acrolein is extremely acrid and irritating to humans when inhaled,  with acute
exposure resulting in upper respiratory tract irritation, mucus hypersecretion and congestion.
Levels considerably lower than 1 ppm (2.3 mg/m3) elicit subjective complaints of eye and
nasal irritation and a decrease in the respiratory rate.121'122 Lesions to the lungs and upper
respiratory tract of rats, rabbits, and hamsters have been observed after subchronic exposure

                                         7-11

-------
Draft Regulatory Impact Analysis

to acrolein. Based on animal data, individuals with compromised respiratory function (e.g.,
emphysema, asthma) are expected to be at increased risk of developing adverse responses to
strong respiratory irritants such as acrolein.  This was demonstrated in mice with allergic
airway-disease by comparison to non-diseased mice in a study of the acute respiratory irritant
effects of acrolein.123

      EPA is currently in the process of conducting an assessment of acute exposure effects
for acrolein. The intense irritancy of this carbonyl has been demonstrated during controlled
tests in human subjects, who suffer intolerable eye and nasal mucosal sensory reactions within
minutes of exposure.124

7.1.1.5.6   Polycyclic Organic Matter (POM)

       POM is generally defined as a large class of organic  compounds which have multiple
benzene rings and a boiling point greater than 100 degrees Celsius. Many of the compounds
included in the class of compounds known as POM are classified by EPA as probable human
carcinogens based on animal data. One of these compounds, naphthalene, is discussed
separately below. Polycyclic aromatic hydrocarbons (PAHs) are a subset of POM that
contain only hydrogen and carbon atoms. A number of PAHs are known or suspected
carcinogens. Recent studies have found that maternal exposures to PAHs (a subclass of
POM) in a population of pregnant women were associated with several adverse birth
outcomes, including low birth weight and reduced length at birth, as well as impaired
cognitive development at age three.125'126 EPA has not yet evaluated these recent studies.

7.1.1.5.7   Naphthalene

        Naphthalene is found in small quantities in gasoline and diesel fuels. Naphthalene
emissions have been measured in larger quantities in both gasoline and diesel exhaust
compared with evaporative emissions from mobile sources, indicating it is primarily a product
of combustion.  EPA released an external review draft of a reassessment of the inhalation
carcinogenicity of naphthalene based on a number of recent animal carcinogenicity studies.127
The draft reassessment completed external peer review.128  Based on external peer review
comments received, additional analyses are being undertaken. This external review draft does
not represent official agency opinion and was released solely for the purposes of external peer
review and public comment.  Once EPA evaluates public and peer reviewer comments,  the
document will be revised. The National Toxicology Program listed naphthalene as
"reasonably anticipated to be a human carcinogen" in 2004 on the basis of bioassays reporting
clear evidence of carcinogenicity  in rats and some evidence of carcinogenicity in mice.129
California EPA has released a new risk assessment for naphthalene, and the IARC has
reevaluated naphthalene and re-classified it as Group 2B: possibly carcinogenic to humans.130
Naphthalene also causes a number of chronic non-cancer effects in animals, including
abnormal cell changes and growth in respiratory and nasal tissues.131

7.1.1.5.8   Other Air Toxics

       In addition to the compounds described above, other compounds in gaseous
hydrocarbon and PM emissions from vehicles would be affected by today's proposed action.

                                        7-12

-------
                                                   Environmental and Health Impacts

Mobile source air toxic compounds that would potentially be impacted include ethylbenzene,
polycyclic organic matter, propionaldehyde, toluene, and xylene. Information regarding the
health effects of these compounds can be found in EPA's IRIS database.m

7.1.2 Environmental Effects Associated with Exposure to Pollutants

       In this section we will discuss the environmental effects associated with non-GHG co-
pollutants, specifically: paniculate matter, ozone, NOx, SOX, carbon monoxide and air toxics.

7.1.2.1  Visibility Degradation

       Emissions from LD vehicles contribute to poor visibility in the U.S. through their
primary PM2 5 and NOx emissions (which contribute to the formation of secondary PM2 5).
These airborne particles degrade visibility by scattering and absorbing light.  Good visibility
increases the quality of life where individuals live and work, and where they engage in
recreational activities.

       The U.S. government places special emphasis on protecting visibility in national parks
and wilderness areas. Section 169 of the Clean Air Act requires the U.S. government to
address existing visibility impairment and future visibility impairment in the 156 national
parks (see Figure 7-1) exceeding 6,000 acres, and wilderness areas exceeding 5,000 acres,
which are categorized as mandatory class I federal  areas (62 FR 38680, July  18, 1997).

7.1.2.1.1   Visibility Monitoring

       In conjunction with the U.S. National Park Service, the U.S. Forest Service, other
Federal land managers, and State organizations in the U.S., the U.S. EPA has supported
visibility monitoring in national parks and wilderness areas since 1988. The monitoring
network was originally  established at 20 sites, but it has now been expanded to 110 sites that
represent all but one of the 156 mandatory Federal Class I areas across the country (see Figure
7-1).  This long-term visibility monitoring network is known as IMPROVE (Interagency
Monitoring of Protected Visual Environments).

       IMPROVE provides direct measurement of fine particles that contribute to visibility
impairment. The IMPROVE network employs aerosol measurements at all sites, and optical
and scene measurements at some of the sites. Aerosol measurements are taken for PMio and
PM2 5 mass, and for key constituents of PM2.s, such as sulfate, nitrate, organic and elemental
carbon, soil dust, and several other elements. Measurements for specific aerosol constituents
are used to calculate "reconstructed" aerosol light extinction by multiplying the mass for each
constituent by its empirically-derived scattering and/or absorption efficiency, with adjustment
for the relative humidity.  Knowledge of the main constituents of a site's light extinction
"budget" is critical for source apportionment and control strategy development.  Optical
measurements are used to directly measure light extinction or its components. Such
measurements are taken principally with either a transmissometer, which measures total light
extinction, or a nephelometer, which measures particle scattering (the largest human-caused
component of total extinction).  Scene characteristics are typically recorded three times daily
with 35 millimeter photography and are used to determine the quality of visibility conditions


                                         7-13

-------
Draft Regulatory Impact Analysis

(such as effects on color and contrast) associated  with specific levels of light extinction as
measured under both direct and aerosol-related methods.  Directly measured light extinction
is used under the IMPROVE protocol to cross check that the aerosol-derived light extinction
levels are reasonable in establishing current visibility conditions. Aerosol-derived light
extinction is used to document spatial and temporal trends and to determine how proposed
changes in atmospheric constituents would affect future visibility conditions.

       Annual average visibility conditions (reflecting light extinction due to both
anthropogenic and non-anthropogenic sources) vary regionally across the U.S. The rural East
generally has higher levels of impairment than remote sites in the West, with the exception of
urban-influenced sites  such as San Gorgonio Wilderness (CA) and Point Reyes National
Seashore (CA), which  have annual average levels comparable to certain sites in the Northeast.
Regional differences are illustrated by Figures 4-3 9a and 4-3 9b in the Air Quality Criteria
Document for Paniculate Matter, which show that, for Class I areas, visibility levels on the
20% haziest days in the West are about equal to levels on the  20% best days in the East.133

       Higher visibility impairment levels in the East are due to generally higher
concentrations of anthropogenic fine particles, particularly sulfates, and higher average
relative humidity levels. In fact, sulfates account for 60-86% of the haziness in eastern
sites.134 Aerosol light extinction due to sulfate on the 20% haziest days is significantly larger
in eastern Class I areas as compared to western areas (Figures 4-40a and 4-40b in the Air
Quality Criteria Document for Paniculate Matter).135  With the exception of remote sites in
the northwestern U.S.,  visibility is typically worse in the summer months.  This is particularly
true in the Appalachian region, where average light extinction in the summer exceeds the
annual average by 40%.136

7.1.2.1.2   Addressing Visibility in the U.S.

        The U.S. EPA is pursuing a two-part strategy to address visibility. First, to address
the welfare effects of PM on visibility, EPA set secondary PM2 5 standards which act in
conjunction with the establishment of a regional haze program.  In setting this secondary
standard, EPA has concluded that PM2 5 causes adverse  effects on visibility in various
locations, depending on PM concentrations and factors such as chemical composition and
average relative humidity. Second, section 169 of the Clean Air Act provides additional
authority to address existing visibility impairment and prevent future visibility impairment in
the 156 national parks, forests and wilderness areas categorized as mandatory Class I federal
areas (62 FR 38680-81, July 18, 1997).: Figure 7-1 below identifies where each of these
parks are located in the U.S. In July 1999, the regional haze rule (64 FR 35714) was put in
place to protect the visibility in mandatory Class I federal areas.  Visibility can be said to be
impaired inbothPM2s nonattainment areas and mandatory Class I federal areas/
1 These areas are defined in section 162 of the Act as those national parks exceeding 6,000 acres, wilderness
areas and memorial parks exceeding 5,000 acres, and all international parks which were in existence on August
7, 1977.
J As mentioned above, the  EPA recently amended the PM NAAQS, making the secondary NAAQS equal, in all
respects, to the primary standards for both PM2.5 and PMi0, (71 FR 61144, Oct. 17, 2006). In February 2009, the
D.C. Circuit Court remanded the secondary standards for fine particles, based on EPA's failure to adequately

                                          7-14

-------
                                                      Environmental and Health Impacts
Produced by NFS Air Resources Division
                             ' Rainbow Lake, Vn and Bradwell Bay, fL are Class 1 Are)
                             where visibility is not an important air quality related value
                         Figure 7-1  Mandatory Class I Areas in the U.S.
7.1.2.2   Plant and Ecosystem Effects of Ozone

         There are a number of environmental or public welfare effects associated with the
presence of ozone in the ambient air.137  In this section we discuss the impact of ozone on
plants, including trees, agronomic crops and urban ornamentals.

       The Air Quality Criteria Document for Ozone and related Photochemical Oxidants
notes that, "ozone affects vegetation throughout the United States, impairing crops, native
vegetation, and ecosystems more than any other air pollutant".138 Like carbon dioxide (CCh)
and other gaseous substances, ozone enters plant tissues primarily through apertures (stomata)
in leaves in a process called "uptake".139 Once sufficient levels of ozone, a highly reactive
substance, (or its reaction products) reaches the interior of plant cells, it can inhibit or damage
essential cellular components and functions, including enzyme activities, lipids, and cellular
membranes, disrupting the plant's osmotic (i.e., water) balance and energy utilization
patterns.140'141  If enough tissue becomes damaged from these effects, a plant's capacity to fix
carbon to form carbohydrates, which are the primary form of energy used by plants is
explain why setting the secondary PM2.5 NAAQS equivalent to the primary standards provided the required
protection for public welfare including protection from visibility impairment.

                                           7-15

-------
Draft Regulatory Impact Analysis

reduced,142 while plant respiration increases. With fewer resources available, the plant
reallocates existing resources away from root growth and storage, above ground growth or
yield, and reproductive processes, toward leaf repair and maintenance, leading to reduced
growth and/or reproduction. Studies have shown that plants stressed in these ways may
exhibit a general loss of vigor, which can lead to secondary impacts that modify plants'
responses to other environmental factors.  Specifically, plants may become more sensitive to
other air pollutants, more susceptible to disease, insect attack, harsh weather (e.g., drought,
frost) and other environmental stresses. Furthermore, there is evidence that ozone can
interfere with the formation of mycorrhiza, essential symbiotic fungi associated with the roots
of most terrestrial plants, by reducing the  amount of carbon available for transfer from the
host to the symbiont.143'144

       This ozone damage  may or may not be accompanied by visible injury on leaves, and
likewise, visible foliar injury may or may not be a symptom of the other types of plant
damage described above. When visible injury is present, it is commonly manifested as
chlorotic or necrotic spots, and/or increased leaf senescence (accelerated leaf aging).  Because
ozone damage can consist of visible injury to leaves, it can also reduce the aesthetic value of
ornamental vegetation and trees in urban landscapes, and negatively affects scenic vistas in
protected natural areas.

       Ozone can produce both acute and chronic injury in sensitive species depending on the
concentration level and the  duration of the exposure. Ozone effects also tend to accumulate
over the growing season of the plant, so that even lower concentrations experienced for a
longer duration have the potential to create chronic stress on sensitive vegetation.  Not all
plants, however, are equally sensitive to ozone. Much of the variation in sensitivity between
individual plants or whole species is related to the plant's ability to regulate the extent of gas
exchange via leaf stomata (e.g.,  avoidance of ozone uptake through closure of
stomata)145'146'147 Other resistance mechanisms may involve the intercellular production of
detoxifying substances. Several biochemical substances capable of detoxifying ozone have
been reported to occur in plants, including the antioxidants ascorbate and glutathione. After
injuries have occurred,  plants may be capable of repairing the damage to a limited extent.148

       Because of the differing sensitivities among plants to ozone, ozone pollution can also
exert a selective pressure that leads to changes in plant community composition. Given the
range of plant sensitivities and the fact that numerous other environmental factors modify
plant uptake and response to ozone, it is not possible to identify threshold values above which
ozone is consistently toxic for all plants. The next few paragraphs present additional
information on ozone damage to trees, ecosystems, agronomic crops and urban ornamentals.

       Ozone also has been conclusively  shown to cause discernible injury to forest
trees.149'150 In terms of forest productivity and ecosystem diversity, ozone may be the
pollutant with the greatest potential for regional-scale forest impacts. Studies have
demonstrated repeatedly that ozone concentrations commonly observed in polluted areas can
have substantial impacts on plant function.151'152

       Because plants are at the base of the food web in many ecosystems, changes to the
plant community can affect associated organisms and ecosystems (including the suitability of

                                          7-16

-------
                                                   Environmental and Health Impacts

habitats that support threatened or endangered species and below ground organisms living in
the root zone). Ozone impacts at the community and ecosystem level vary widely depending
upon numerous factors, including concentration and temporal variation of tropospheric ozone,
species composition, soil properties and climatic factors.153  In most instances, responses to
chronic or recurrent exposure in forested ecosystems are subtle and not observable for many
years.  These injuries can cause stand-level forest decline in sensitive ecosystems.154'155'156  It
is not yet possible to predict ecosystem responses to ozone with much certainty; however,
considerable knowledge of potential ecosystem responses has been acquired through long-
term observations in highly damaged forests in the United States.

       Laboratory and field experiments have also shown reductions in yields for agronomic
crops exposed to ozone, including vegetables (e.g., lettuce) and field crops (e.g., cotton and
wheat). The most extensive field experiments, conducted under the National Crop Loss
Assessment Network (NCLAN) examined 15 species and numerous cultivars.  The NCLAN
results show that "several economically important crop species are sensitive to ozone levels
typical of those found in the United States."157 In addition, economic studies have shown
reduced economic benefits as a result of predicted reductions in crop yields associated with
observed ozone levels.158'159'160

       Urban ornamentals represent an additional vegetation category likely to experience
some degree of negative effects associated with exposure to ambient ozone levels.  It is
estimated that more than $20 billion (1990 dollars) are spent annually on landscaping using
ornamentals, both by private property owners/tenants and by governmental units responsible
for public areas.161 This is therefore a potentially costly environmental effect.  However, in
the absence of adequate exposure-response functions and economic damage functions for the
potential range of effects relevant to these types of vegetation, no direct quantitative analysis
has been conducted.

       Air pollution can have noteworthy cumulative impacts on forested ecosystems by
affecting regeneration, productivity, and species composition.162 In the U.S., ozone in the
lower atmosphere is one of the pollutants of primary  concern.  Ozone injury to forest plants
can be diagnosed by examination of plant leaves.  Foliar injury is usually the first visible sign
of injury to plants from ozone exposure and indicates impaired physiological processes in the
leaves.163

       In the U.S. this indicator is based on data from the U.S. Department of Agriculture
(USD A) Forest Service Forest Inventory and Analysis (FIA) program.  As part of its Phase 3
program, formerly known as Forest Health Monitoring, FIA examines ozone injury to ozone-
sensitive plant species at ground monitoring sites in forest land across the country. For this
indicator, forest land does not include woodlots and urban trees. Sites are selected using a
systematic sampling grid, based on a global sampling design.164'165  At each site that has at
least 30 individual plants of at least three ozone-sensitive species and enough open space to
ensure that sensitive plants are not protected from ozone exposure by the forest canopy, FIA
looks for damage on the foliage of ozone-sensitive forest plant species. Monitoring of ozone
injury to plants by the USDA Forest Service has expanded over the last 10 years from
monitoring sites in 10 states in 1994 to nearly 1,000 monitoring sites in 41 states in 2002.
                                         7-17

-------
Draft Regulatory Impact Analysis
7.1.2.2.1   Recent Ozone Data for the U.S.

        There is considerable regional variation in ozone-related visible foliar injury to
sensitive plants in the U.S. The U.S. EPA has developed an environmental indicator based on
data from the U.S. Department of Agriculture (USDA) Forest Service Forest Inventory and
Analysis (FIA) program which examines ozone injury to ozone-sensitive plant species at
ground monitoring sites in forest land across the country (This indicator does not include
woodlots and urban trees). Sites are selected using a systematic sampling grid, based on a
global sampling design.166'167 Because ozone injury is cumulative over the course of the
growing season, examinations are conducted in July and August, when ozone injury is
typically highest.  The data underlying the indictor in Figure 7-2 are based on averages of all
observations collected in 2002, the latest year for which data are publicly available at the time
the study was conducted, and are broken down by U.S. EPA Region. Ozone damage to forest
plants is classified using a subjective five-category biosite index based on expert opinion, but
designed to be equivalent from site to site. Ranges of biosite values translate to no injury, low
or moderate foliar injury (visible foliar injury to highly sensitive or moderately sensitive
plants, respectively), and high or severe  foliar injury, which would be expected to result in
tree-level or ecosystem-level responses,  respectively.168'169

        The highest percentages of observed high and severe foliar injury,  those which are
most likely to be associated with tree or  ecosystem-level responses, are primarily found in the
Mid-Atlantic and Southeast regions.  In EPA Region 3 (which comprises the States of
Pennsylvania, West Virginia, Virginia, Delaware, Maryland and Washington D.C.), 12% of
ozone-sensitive plants showed signs of high or severe foliar damage, and in Regions 2 (States
of New  York, New Jersey), and 4 (States of North Carolina, South Carolina, Kentucky,
Tennessee, Georgia, Florida, Alabama, and Mississippi) the values were 10% and 7%,
respectively.  The sum of high and severe ozone injury ranged from 2% to  4% in EPA Region
1 (the six New England States), Region 7 (States of Missouri, Iowa, Nebraska and Kansas),
and Region 9 (States of California, Nevada, Hawaii and Arizona). The percentage of sites
showing some ozone damage was about 45% in each of these EPA Regions.
                                        7-18

-------
                                                       Environmental and Health Impacts
                             Degree of injury:
                               None    Low   Moderate  High   Severe
                             Percent a! monitoring sites in each category:
                    Region 1
                    (54 sites)
&S.5
16.7
11.1
-3.7
Region 2
(42 sites'!
Region 3
•1T UK'
Region 4
;227iies;
61.9
21.4
7.1

55.9
1B.Q

75.3
14.4

10.1
7 ^
7.1

7.2


L4
-.5
O.E
^:
                    Region 5  [
                    :isc ;;«:!
 75.6
   1B3
Region 6
(59 shss'i
Region 7
(63 sites)
94.9
-

S5.7
9.5

::'
•32
•1 5
                    Regicfi 8
                    (72 sites)

                    Region 9
                    (30 sites)

                    Region 10
                    (57 sites)
      100.0
 76.3
  12.5
      too.o
                    'Cmwraga: 945 monitoring sites,
                    tacated in 41 states.
                    :Totals may not add to 100% dut to
                    rounding
                    OjlB saures: LfS£M fargsf Sen/ice,
                    2006
               EPA Regions
                       of.


             Figure 7-2 Ozone Injury to Forest Plants in U.S. by EPA Regions, 2002al
        7.1.2.2.1.1    Indicator Limitations

       Field and laboratory studies were reviewed to identify the forest plant species in each
region that are highly sensitive to ozone air pollution. Other forest plant species, or even
                                           7-19

-------
Draft Regulatory Impact Analysis

  genetic variants of the same species, may not be harmed at ozone levels that cause effects on
  the selected ozone-sensitive species.

       Because species distributions vary regionally, different ozone-sensitive plant species
  were examined in different parts of the country. These target species could vary with
  respect to ozone sensitivity, which might account for some of the apparent differences in
  ozone injury among regions of the U.S.

       Ozone damage to foliage is considerably reduced under conditions of low soil
  moisture, but most of the variability in the index (70%) was explained by ozone
  concentration.170 Ozone may have other adverse impacts on plants (e.g., reduced
  productivity) that do not show signs of visible foliar injury.171

       Though FIA has extensive spatial coverage based on a robust sample design, not all
  forested areas in the U.S. are monitored for ozone injury.  Even though the biosite data have
  been collected over multiple years, most biosites were not monitored over the entire period,
  so these data cannot provide more than a baseline for future trends.

7.1.2.3   Ozone Impacts on Forest Health

        Air pollution can impact the environment and affect ecological systems, leading to
changes in the biological community (both in the diversity of species and the health and vigor
of individual species).  As an example, many studies have shown that ground-level ozone
reduces the health of plants including many commercial and ecologically important forest tree
species throughout the United States.172

       When ozone is present in the air, it can enter the leaves of plants, where it can cause
significant cellular damage.  Since photosynthesis occurs in cells within leaves, the ability of
the plant to produce energy by photosynthesis can be compromised if enough damage occurs
to these cells.  If enough tissue becomes damaged it can reduce carbon fixation and increase
plant respiration, leading to reduced growth and/or reproduction in young and mature trees.
Ozone stress also increases the susceptibility of plants to disease, insects, fungus, and other
environmental stressors (e.g., harsh weather). Because ozone damage can consist of visible
injury to leaves, it also  reduces the aesthetic value of ornamental vegetation and trees in urban
landscapes, and negatively affects scenic vistas in protected natural areas.

       Assessing the impact of ground-level ozone on forests in the eastern United States
involves understanding the risks to sensitive tree species from ambient ozone concentrations
and accounting for the prevalence of those species within the forest.  As a way to quantify the
risks to particular plants from ground-level ozone, scientists have developed ozone-
exposure/tree-response functions by exposing tree seedlings to different ozone levels and
measuring reductions in growth as "biomass loss." Typically, seedlings are used because they
are easy to manipulate and measure their growth loss from ozone pollution.  The mechanisms
of susceptibility to ozone  within the leaves of seedlings and mature trees are identical, though
the magnitude of the effect may be higher or lower depending on the tree species. 173
                                         7-20

-------
                                                    Environmental and Health Impacts

Some of the common tree species in the United States that are sensitive to ozone are black
cherry (Primus serotina), tulip-poplar (Liriodendron tulipifera), eastern white pine (Pinus
strobus). Ozone-exposure/tree-response functions have been developed for each of these tree
species, as well as for aspen (Populus tremuliodes), and ponderosa pine (Pinusponderoso).
Other common tree species, such as oak (Quercus spp.) and hickory (Carya spp.), are not
nearly as sensitive to ozone. Consequently, with knowledge of the distribution of sensitive
species and the level of ozone at particular locations, it is possible to estimate a "biomass
loss" for each species across their range.

7.1.2.4   Participate Matter Deposition

       Paniculate matter contributes to adverse effects on vegetation and ecosystems, and to
soiling and materials damage. These welfare effects result predominately from exposure to
excess amounts of specific chemical species, regardless of their source or predominant form
(particle, gas or liquid).  Reflecting this fact, the PM AQCD concludes that regardless of size
fractions, particles containing nitrates and sulfates have the greatest potential for widespread
environmental significance, while effects are also related to other chemical constituents found
in ambient PM, such as trace metals and organics. The following characterizations of the
nature of these welfare effects are based on the information contained in the PM AQCD and
PM Staff Paper.174'175

7.1.2.4.1   Deposition of Nitrogen and Sulfur

        Nitrogen and sulfur interactions in the environment are highly  complex.  Both are
essential, and sometimes limiting, nutrients needed for growth and productivity. Excesses of
nitrogen or sulfur can lead to soil and water acidification, nutrient enrichment, and
eutrophication.176

       The process of acidification affects both freshwater aquatic and  terrestrial ecosystems.
Acid deposition causes acidification of sensitive surface waters. The effects of acid deposition
on aquatic systems depend largely upon the ability of the ecosystem to neutralize the
additional acid. As acidity increases, aluminum leached from soils and sediments, flows into
lakes and streams and can be toxic to both terrestrial and aquatic biota.  The lower pH
concentrations and higher aluminum levels resulting from acidification make it difficult for
some fish and other aquatic organisms to survive, grow, and reproduce. Research on effects
of acid deposition on forest ecosystems has come to focus increasingly  on the biogeochemical
processes that affect uptake, retention, and cycling of nutrients within these  ecosystems.
Decreases in available base cations from soils are at least partly attributable to acid
deposition.  Base  cation depletion is a cause for concern because of the  role these ions play in
acid neutralization and, because calcium, magnesium and potassium are essential nutrients for
plant growth and physiology.  Changes in the relative proportions of these nutrients,
especially in comparison with aluminum concentrations, have been associated with declining
forest health.

       At current ambient levels, risks to vegetation from short-term exposures to dry
deposited paniculate nitrate or sulfate are low. However, when found in acid or acidifying
deposition,  such particles do have the potential to cause direct leaf injury. Specifically, the

                                         7-21

-------
Draft Regulatory Impact Analysis

responses of forest trees to acid precipitation (rain, snow) include accelerated weathering of
leaf cuticular surfaces, increased permeability of leaf surfaces to toxic materials, water, and
disease agents; increased leaching of nutrients from foliage; and altered reproductive
processes—all which serve to weaken trees so that they are more susceptible to other stresses
(e.g., extreme weather, pests, pathogens). Acid deposition with levels of acidity associated
with the leaf effects described above are currently found in some locations in the eastern
U.S.177 Even higher concentrations of acidity can be present in occult depositions (e.g., fog,
mist or clouds) which more frequently impacts higher elevations. Thus,  the risk of leaf injury
occurring from acid deposition in some areas of the eastern U.S. is high. Nitrogen deposition
has also been shown to impact ecosystems in the western U.S.  A study conducted in the
Columbia River Gorge National Scenic Area (CRGNSA), located along a portion of the
Oregon/Washington border,  indicates that lichen communities  in the CRGNSA have shifted
to a higher proportion of nitrophilous species and the nitrogen  content of lichen tissue is
elevated.178 Lichens are sensitive indicators of nitrogen deposition effects to terrestrial
ecosystems and the lichen studies in the Columbia River Gorge clearly show that ecological
effects from air pollution are occurring.

       Some of the  most significant detrimental effects associated with excess nitrogen
deposition are those associated with a syndrome known as nitrogen saturation. These effects
include: (1) decreased productivity, increased mortality, and/or shifts in plant community
composition,  often leading to decreased biodiversity in many natural habitats wherever
atmospheric reactive nitrogen deposition increases significantly above background and critical
thresholds are exceeded; (2)  leaching of excess nitrate and associated base cations from soils
into streams, lakes, and rivers, and mobilization of soil aluminum; and (3)  fluctuation of
ecosystem processes such as nutrient and energy cycles through changes in the functioning
and species composition of beneficial soil organisms.179

       In the U.S. numerous forests now show severe symptoms of nitrogen saturation.
These forests include: the northern hardwoods and mixed conifer forests in the Adirondack
and Catskill Mountains of New York; the red spruce forests at  Whitetop Mountain, Virginia,
and Great Smoky Mountains National Park, North Carolina; mixed hardwood watersheds at
Fernow Experimental Forest in West Virginia; American beech forests in Great Smoky
Mountains National Park, Tennessee; mixed conifer forests and chaparral watersheds in
southern California and the southwestern Sierra Nevada in Central California; the alpine
tundra/subalpine conifer forests of the Colorado Front Range; and red alder forests in the
Cascade Mountains in Washington.

       Excess nutrient inputs into aquatic ecosystems (i.e. streams, rivers, lakes, estuaries or
oceans) either from direct atmospheric deposition, surface runoff, or leaching from nitrogen
saturated soils into ground or surface waters can contribute  to conditions of severe water
oxygen depletion; eutrophication and algae blooms; altered fish distributions, catches, and
physiological states; loss of biodiversity; habitat degradation; and increases in the incidence
of disease.

       Atmospheric deposition of nitrogen is a significant source of total nitrogen to many
estuaries in the United States. The amount of nitrogen entering estuaries that is ultimately
attributable to atmospheric deposition is not well-defined. On an annual basis, atmospheric

                                         7-22

-------
                                                   Environmental and Health Impacts

nitrogen deposition may contribute significantly to the total nitrogen load, depending on the
size and location of the watershed. In addition, episodic nitrogen inputs, which may be
ecologically important, may play a more important role than indicated by the annual average
concentrations. Estuaries in the U.S. that suffer from nitrogen enrichment often experience a
condition known as eutrophication. Symptoms of eutrophication include changes in the
dominant species of phytoplankton, low levels of oxygen in the water column, fish and
shellfish kills, outbreaks of toxic alga, and other population changes which can cascade
throughout the food web. In addition, increased phytoplankton growth in the water column
and on surfaces can attenuate light causing declines in submerged aquatic vegetation, which
serves as an important habitat for many estuarine fish and shellfish species.

       Severe and persistent eutrophication often directly impacts human activities.  For
example, losses in the nation's fishery resources may be directly caused by fish kills
associated with low dissolved oxygen and toxic blooms. Declines in tourism occur when low
dissolved oxygen causes noxious smells and floating mats of algal blooms create unfavorable
aesthetic conditions.  Risks to human health increase when the toxins from algal blooms
accumulate in edible fish and shellfish, and when toxins become airborne, causing respiratory
problems due to inhalation. According to a NO A A report, more than half of the nation's
estuaries have moderate to high expressions of at least one of these symptoms - an indication
that eutrophication is well developed in more than half of U.S. estuaries.180

7.1.2.4.2  Deposition  of Heavy Metals

        Heavy metals, including cadmium, copper, lead, chromium, mercury, nickel and
zinc, have the greatest potential for influencing forest growth (PM AQCD, p. 4-87).1S1
Investigation of trace metals near roadways and industrial facilities indicate that a substantial
load of heavy metals can accumulate on vegetative surfaces. Copper, zinc, and nickel have
been documented to cause direct toxicity to vegetation under field conditions (PM AQCD, p.
4-75). Little research has been conducted on the  effects associated with mixtures of
contaminants found in ambient PM.  While metals typically exhibit low solubility, limiting
their bioavailability and direct toxicity, chemical transformations of metal compounds occur
in the environment, particularly in the presence of acidic or other oxidizing species.  These
chemical changes influence the mobility and toxicity of metals in the environment. Once
taken up into plant tissue, a metal compound can undergo chemical changes, accumulate and
be passed along to  herbivores or can re-enter the soil and further cycle in the environment.
Although there has been no direct evidence of a physiological association between tree injury
and heavy metal exposures, heavy  metals have been implicated because of similarities
between metal deposition patterns  and forest decline (PM AQCD, p. 4-76).  This
hypothesized relationship/correlation was further explored in high elevation forests in the
northeastern U.S.  These studies measured levels of a group of intracellular compounds found
in plants that bind with metals and are produced by plants as a response to sublethal
concentrations of heavy metals. These studies indicated a systematic and significant increase
in concentrations of these compounds associated with the extent of tree injury.  These data
strongly imply that metal stress causes tree injury and contributes to forest decline in the
northeastern United States (PM AQCD 4-76,77).182  Contamination of plant leaves by heavy
metals can lead to elevated soil levels.  Trace metals absorbed into the plant frequently bind to
the leaf tissue, and then are lost when the leaf drops (PM AQCD, p. 4-75). As the fallen

                                         7-23

-------
Draft Regulatory Impact Analysis

leaves decompose, the heavy metals are transferred into the soil.183'184

       The environmental sources and cycling of mercury are currently of particular concern
due to the bioaccumulation and biomagnification of this metal in aquatic ecosystems and the
potent toxic nature of mercury in the forms in which is it ingested by people and other
animals.  Mercury is unusual compared with other metals in that it largely partitions into the
gas phase (in elemental form), and therefore has a longer residence time in the atmosphere
than a metal found predominantly in the particle phase. This property enables mercury to
travel far from the primary source before being deposited and accumulating in the aquatic
ecosystem.  The major source of mercury in the Great Lakes is from atmospheric deposition,
accounting  for approximately eighty percent of the mercury in Lake Michigan.185'186 Over
fifty percent of the mercury in the Chesapeake Bay has been attributed to atmospheric
deposition.187 Overall, the National Science and Technology Council identifies atmospheric
deposition as the primary source of mercury to aquatic systems.188 Forty-four states have
issued health advisories for the consumption of fish contaminated by mercury; however, most
of these advisories are issued in areas without a mercury point source.

       Elevated levels of zinc and lead have been identified in streambed sediments, and
these elevated levels have been correlated with population density  and motor vehicle
use.189'190 Zinc and nickel have also been identified in urban water and soils.  In addition,
platinum, palladium, and rhodium, metals found in the catalysts of modern motor vehicles,
have been measured at elevated levels along roadsides.191  Plant uptake of platinum has been
observed at these locations.

7.1.2.4.3    Deposition ofPolycyclic Organic Matter

       Poly cyclic organic matter (POM) is a byproduct of incomplete combustion and
consists of organic compounds with more than one benzene ring and a boiling point greater
than or equal to 100 degrees centigrade.192 Polycyclic aromatic hydrocarbons (PAHs) are a
class of POM that contains compounds which are known or suspected carcinogens.

       Major sources of PAHs include mobile sources. PAHs in the environment may be
present as a gas or adsorbed onto airborne paniculate matter. Since the majority of PAHs are
adsorbed onto particles less than 1.0 um in diameter, long range transport is possible.
However, studies have shown that PAH compounds adsorbed onto diesel exhaust paniculate
and exposed to ozone have half lives of 0.5 to 1.0 hours.193

       Since PAHs are insoluble, the compounds generally are particle reactive and
accumulate in sediments.  Atmospheric deposition of particles is believed to be the major
source of PAHs to the sediments of Lake Michigan.194'195  Analyses of PAH deposition in
Chesapeake and Galveston Bay indicate that dry deposition and gas exchange from the
atmosphere to the surface water predominate.196'197 Sediment concentrations of PAHs are
high enough in some segments of Tampa Bay to pose  an environmental health threat. EPA
funded a study to better characterize the sources and loading rates for PAHs into Tampa
Bay.198 PAHs  that enter a water body through gas exchange likely partition into organic rich
particles and can be biologically recycled, while dry deposition of aerosols containing PAHs
tend to be more resistant to biological recycling.199 Thus, dry deposition is likely the main

                                        7-24

-------
                                                   Environmental and Health Impacts

pathway for PAH concentrations in sediments while gas/water exchange at the surface may
lead to PAH distribution into the food web, leading to increased health risk concerns.

       Trends in PAH deposition levels are difficult to discern because of highly variable
ambient air concentrations, lack of consistency in monitoring methods, and the significant
influence of local sources on deposition levels.200 Van Metre et al. noted PAH concentrations
in urban reservoir sediments have increased by 200-300% over the last forty years and
correlate with increases in automobile use.201

       Cousins et al. estimate that more than ninety percent of semi-volatile organic
compound (SVOC) emissions in the United Kingdom deposit on soil.202 An analysis of PAH
concentrations near a Czechoslovakian roadway indicated that concentrations were thirty
times greater than background.203

7.1.2.4.4  Materials Damage and Soiling

        The effects of the deposition of atmospheric pollution, including ambient PM, on
materials are related to both physical damage and impaired aesthetic qualities.  The deposition
of PM  (especially sulfates and nitrates) can physically affect materials, adding to the effects of
natural weathering processes, by potentially promoting or accelerating the corrosion of
metals, by degrading paints, and by deteriorating building materials such as concrete and
limestone. Only chemically active fine particles or hygroscopic coarse particles contribute to
these physical effects. In addition, the deposition of ambient PM can reduce the aesthetic
appeal of buildings and culturally important articles through soiling. Particles consisting
primarily of carbonaceous compounds cause soiling of commonly used building materials and
culturally important items such as statues and works of art.

7.1.2.5   Environmental Effects of Air Toxics

        Fuel combustion emissions contribute to  ambient levels of pollutants that contribute
to adverse effects on vegetation. Volatile organic compounds (VOCs), some of which are
considered air toxics, have long been suspected to play a role in vegetation damage.204 In
laboratory experiments, a wide range of tolerance to VOCs has been observed.205 Decreases
in harvested seed pod weight have been reported  for the more sensitive plants, and some
studies have reported effects on seed germination, flowering and fruit  ripening. Effects of
individual VOCs or their role  in conjunction with other stressors (e.g., acidification, drought,
temperature extremes) have not been well studied.  In a recent study of a mixture of VOCs
including ethanol and toluene on herbaceous plants, significant effects on seed production,
leaf water content and photosynthetic efficiency were reported for some  plant species.206

       Research suggests an adverse impact of vehicle exhaust on plants, which has in some
cases been attributed to aromatic compounds and in other cases to nitrogen oxides.207'208'209
The impacts of VOCs on plant reproduction may have long-term implications for biodiversity
and survival of native species near major roadways. Most of the studies of the impacts of
VOCs  on vegetation have focused on short-term exposure and few studies have focused on
long-term effects of VOCs on vegetation and the  potential for metabolites of these compounds
to affect herbivores or insects.
                                         7-25

-------
Draft Regulatory Impact Analysis



7.2  Non-GHG Air Quality Impacts

7.2.1 Introduction

       Chapter 5 of this DRIA presents the projected emissions changes due to the proposed
rule. Once the emissions changes are projected the next step is to look at how the ambient air
quality would be impacted by those emissions changes.  Although the purpose of this proposal
is to address greenhouse gas emissions, this proposed rule would also impact emissions of
criteria and hazardous air pollutants.  Section 7.2.2 describes current ambient levels of PM,
ozone, CO and some air toxics without the standards being proposed in this rule. No air
quality modeling was done for this DRIA to project the impacts of the proposed rule. Air
quality modeling will be done for the final rule, however, and those plans are discussed in
Section 7.2.3.

7.2.2  Current Levels of Pollutants

7.2.2.1   Participate Matter

       As described in Section 7.1.1.1, PM causes adverse health effects, and the U.S.
government has set national standards to provide requisite protection against those health
effects. There are two U.S. national ambient air quality standards (NAAQS) for PM2 5: an
annual standard (15 ug/m3) and a 24-hour standard (35 ug/m3). The most recent revisions to
these standards were in 1997  and 2006. In 2005 the U.S. EPA designated nonattainment areas
for the 1997  PM25 NAAQS (70 FR 19844, April 14, 2005).K  As of June 5, 2009 there are 39
1997 PM2 5 nonattainment areas comprised of 208 full or partial counties with a total
population exceeding 88 million. Area designations for the 2006 24-hour PM2 5 NAAQS are
expected to be promulgated in 2009 and become effective 90 days after publication in the
Federal Register.

       States with PM2 5 nonattainment areas will be required to take action to bring those
areas into compliance in the future. Most 1997 PM2 5 nonattainment areas are required to
attain the 1997 PM25 NAAQS in the  2010 to 2015 time frame and then be required to
maintain the 1997 PM2 5 NAAQS thereafter.210 The 2006  24-hour PM2 5 nonattainment areas
will be required to attain the 2006 24-hour PM2 5 NAAQS in the 2014 to  2019 time frame and
then be required to  maintain the 2006 24-hour PM25 NAAQS thereafter.211
7.2.2.2   Ozone

        As described in Section 7.1.1.2, ozone causes adverse health effects, and the U.S.
government has set national standards to protect against those health effects.  The NAAQS for
ozone is an 8-hour standard set at 0.075 ppm.  The most recent revision to this standard was in
K A nonattainment area is defined in the Clean Air Act (CAA) as an area that is violating an ambient standard or
is contributing to a nearby area that is violating the standard.

                                        7-26

-------
                                                      Environmental and Health Impacts

2008; the previous 8-hour ozone standard, set in 1997, had been 0.08 ppm. In 2004 the U.S.
EPA designated nonattainment areas for the 1997 8-hour ozone NAAQS (69 FR 23858, April
30, 2004).L As of June 5, 2009  there are 55 1997 8-hour ozone nonattainment areas
comprised of 290  full or partial  counties with a total population of approximately 132
million.212 Nonattainment designations for the 2008 8-hour ozone standard are currently
under development.

       States with ozone nonattainment areas are required to take action to bring those areas
into compliance in the future. The attainment date assigned to an ozone nonattainment area is
based on the area's classification.  Most ozone nonattainment areas are required to attain the
1997 8-hour ozone NAAQS in the 2007 to 2013 time frame and then to maintain it
thereafter.M  The attainment dates associated with the potential nonattainment areas based on
the 2008 8-hour ozone NAAQS will likely be in the 2013 to 2021 timeframe, depending on
the severity of the problem in each area. Table 7-2 provides an estimate, based on 2004-06 air
quality data, of the counties with design values greater than the 2008 ozone NAAQS.

  Table 7-2 Counties with Design Values Greater Than the 2008 Ozone NAAQS Based on 2005-2007 Air
                                       Quality Data

1997 Ozone Standard: counties within the 57
areas currently designated as nonattainment
2008 Ozone Standard: additional counties that
would not meet the 2008 NAAQSb
Total
NUMBER OF
COUNTIES
293
227
520
POPULATIONA
131,977,890
41,285,262
173,263,152
Notes:
a Population numbers are from 2000 census data.
b Attainment designations for the 2008 ozone NAAQS have not yet been made.  Nonattainment for the 2008
Ozone NAAQS will be based on three years of air quality data from later years.  Also, the county numbers in the
table include only the counties with monitors violating the 2008 Ozone NAAQS. The numbers in this table may
be an underestimate of the number of counties and populations that will eventually be included in areas with
multiple counties designated nonattainment.
7.2.2.3    Carbon Monoxide

         As described in Section 7.1.1.4, CO causes adverse health effects, and the U.S.
government has set national standards to protect against those health effects.  There are two
CO NAAQS.  The 8-hour average CO NAAQS is 9 ppm, not to be exceeded more than once
per year, and the 1-hour average CO NAAQS is 35 ppm, not to be exceeded more than once
L A nonattainment area is defined in the Clean Air Act (CAA) as an area that is violating an ambient standard or
is contributing to a nearby area that is violating the standard.
M The Los Angeles South Coast Air Basin 8-hour ozone nonattainment area is designated as severe and will have
to attain before June 15, 2021. The South Coast Air Basin has requested to be reclassified as an extreme
nonattainment area which will make its attainment date June 15, 2024. The San Joaquin Valley Air Basin 8-hour
ozone nonattainment area is designated as serious and will have to attain before June 15, 2013. The San Joaquin
Valley Air Basin has requested to be reclassified as an extreme nonattainment area which will make its
attainment date June 15, 2024.
                                           7-27

-------
Draft Regulatory Impact Analysis

per year. The two standards are the identical standards that were promulgated in 1971.
Reviews of the CO NAAQS in both 1985 and 1994 kept the standards at the same levels. As
of June 5, 2009 there are approximately 479,000 people living in a portion of Clark Co., NV
which is currently the only area in the country that is designated as nonattainment for CO, see
Table 7-3,213  The CO NAAQS is currently under review and a final rule is expected to be
final in May 2011.

            Table 7-3 Classified Carbon Monoxide Nonattainment Areas as of June 2009
AREA
Clark County
(p), NV
Total
CLASSIFICATION
Nonattainment

POPULATION
478,766
478,766
7.2.2.4   Air Toxics

         According to the National Air Toxics Assessment (NATA) for 2002, mobile sources
were responsible for 47 percent of outdoor toxic emissions and over 50 percent of the cancer
risk.214 Nearly the entire U.S. population was exposed to an average concentration of air
toxics that has the potential for adverse noncancer respiratory health effects. EPA recently
finalized vehicle and fuel controls to reduce mobile source air toxics.215 In addition, over the
years, EPA has implemented a number of mobile source and fuel controls resulting in VOC
reductions, which also reduce air toxic emissions.  Modeling from the recent Mobile Source
Air Toxics (MS AT) rule suggests that the mobile source contribution to ambient benzene
concentrations is projected to decrease over 40% by 2015, with a decrease in ambient benzene
concentration from all sources of about 25%.  Although benzene is used as an example, the
downward trend is projected for other air toxics as well. See the RIA for the final MS AT rule
for more information on ambient air toxics projections.216

7.2.3 Impacts on Future Air Quality

        Air quality models use mathematical and numerical techniques to simulate  the
physical and chemical processes that affect air pollutants as they disperse  and react in the
atmosphere. Based on inputs of meteorological data and source information, these models are
designed to characterize primary pollutants that are emitted directly into the atmosphere and
secondary pollutants that are formed as a result of complex chemical reactions within the
atmosphere. Photochemical air quality models have become widely recognized and routinely
utilized tools for regulatory analysis by assessing the effectiveness of control strategies.
These models are applied at multiple spatial scales from local, regional, national, and global.
Section 7.2.3.1 provides more detail on the photochemical model, the Community Multi-scale
Air Quality (CMAQ) model, which will be utilized for the final rule analysis.
                                        7-28

-------
                                                   Environmental and Health Impacts

7.2.3.1   Community Multi-scale Air Quality (CMAQ) Modeling Plans

       Full-scale photochemical air quality modeling is necessary to accurately project levels
of PM2.5, ozone, CO and air toxics.  For the final rule, a national-scale air quality modeling
analysis will be performed to analyze the impacts of the vehicle standards on PM2 5, ozone,
and selected air toxics (i.e., benzene, formaldehyde, acetaldehyde, acroleinand 1,3-
butadiene). The length of time needed to prepare the necessary emissions inventories, in
addition to the processing time associated with the modeling itself, has precluded us from
performing air quality modeling for this proposal.

       Section II.G.l of the preamble presents projections of the changes in criteria pollutant
and air toxics emissions due to the proposed vehicle standards; the basis for those estimates is
set out in Chapter 5 of the DRIA. The atmospheric chemistry related to ambient
concentrations of PM2 5, ozone and air toxics is very complex, and making predictions based
solely on emissions changes is extremely difficult. However, based on the magnitude of the
emissions changes predicted to result from the proposed vehicle standards, we expect that
there will be an improvement in ambient air quality, pending a more comprehensive analysis
for the final rule.

       For the final rule, EPA intends to use a 2005-based Community Multi-scale Air
Quality (CMAQ) modeling platform as the tool for the  air quality modeling. The  CMAQ
modeling system is a comprehensive three-dimensional grid-based Eulerian air quality model
designed to estimate the formation and fate of oxidant precursors, primary and secondary PM
concentrations and deposition, and air toxics, over regional and urban spatial scales (e.g., over
the contiguous U.S.).2 7'218-219  The CMAQ model is a well-known and well-established tool
and is commonly used by EPA for regulatory analyses, for instance the recent ozone NAAQS
proposal, and by States in developing attainment demonstrations for their State
Implementation Plans.220  The CMAQ model (version 4.6) was peer-reviewed in February of
2007 for EPA as reported in "Third  Peer Review of CMAQ Model," and the EPA Office of
Research and Development (ORD) peer review report which includes version 4.7  is currently
being finalized.221

       CMAQ includes many science modules that simulate the emission, production, decay,
 deposition and transport of organic and inorganic gas-phase and particle-phase pollutants in
 the atmosphere. We intend to use the most recent CMAQ version (version 4.7), which was
 officially released by EPA's Office of Research and Development (ORD) in December 2008
 and reflects updates to earlier versions in a number of areas to improve the underlying
 science. These include (1) enhanced secondary organic aerosol (SOA) mechanism to include
 chemistry of isoprene, sesquiterpene, and aged in-cloud biogenic SOA in addition to terpene;
 (2) improved vertical convective mixing; (3) improved heterogeneous reaction involving
 nitrate formation; and (4) an updated gas-phase chemistry mechanism, Carbon Bond 05
 (CB05), with extensions to model explicit concentrations of air toxic species as well as
 chlorine and mercury. This mechanism, CB05-toxics, also computes concentrations of
 species that are involved in aqueous chemistry and that are precursors to aerosols.

       The CMAQ modeling domain will encompass all of the lower 48 States and portions
of Canada and Mexico. The modeling domain will include a large continental U.S. 36 km

                                        7-29

-------
Draft Regulatory Impact Analysis

grid and two 12 km grids (an Eastern US and a Western US domain), as shown in Figure 7-3.
The modeling domain will contain 14 vertical layers with the top of the modeling domain at
about 16,200 meters, or 100 millibars (mb).
                 1Stan WMm Domain rvWMPi
                  On -2-H2000 -972X0
                  fjt3 w* 192

                Figure 7-3  CMAQ 12-km Eastern and Western US modeling domains
       The key inputs to the CMAQ model include emissions from anthropogenic and
biogenic sources, meteorological data, and initial and boundary conditions.  The CMAQ
meteorological input files will be derived from simulations of the Pennsylvania State
University / National Center for Atmospheric Research Mesoscale Model222 for the entire
year of 2005.  This model, commonly referred to as MM5, is a limited-area, nonhydrostatic,
terrain-following system that solves for the full set of physical and thermodynamic equations
which govern atmospheric motions.223 The meteorology for the national 36 km grid and the
12 km Eastern and Western U.S. grids will be developed by EPA and described in more detail
within the final RIA and the technical support document for the final rule air quality
modeling.
       The lateral boundary and initial species concentrations will be provided by a three-
dimensional global atmospheric chemistry model, the GEOS-CHEM model.224 The global
GEOS-CHEM model simulates atmospheric chemical and physical processes driven by
assimilated meteorological observations from the NASA's Goddard Earth Observing System
                                       7-30

-------
                                                   Environmental and Health Impacts

(GEOS). This model will be ran for 2005 with a grid resolution of 2 degree x 2.5 degree
(latitude-longitude) and 20 vertical layers.  The predictions will be used to provide one-way
dynamic boundary conditions at three-hour intervals and an initial concentration field for the
36 km CMAQ simulations.  The future base conditions from the 36 km coarse grid modeling
will be used as the initial/boundary state for all subsequent 12 km finer grid modeling.
7.3 Quantified and Monetized Co-Pollutant Health and Environmental
    Impacts

        This section presents EPA's analysis of the co-pollutant health and environmental
impacts that can be expected to occur as a result of the proposed light-duty vehicle GHG rale.
GHG emissions are predominantly the byproduct of fossil fuel combustion processes that also
produce criteria and hazardous air pollutants. The vehicles that are subject to the proposed
standards are also significant sources of mobile source air pollution such as direct PM, NOx,
VOCs and air toxics. The proposed standards would affect exhaust emissions of these
pollutants from vehicles. They would also affect emissions from upstream sources related to
changes in fuel consumption. Changes in ambient ozone, PM2 5, and air toxics that would
result from the proposed standards are expected to affect human health in the form of
premature deaths and other serious human health effects, as well as other important public
health and welfare effects.

       It is important to quantify the health and environmental impacts associated with the
proposed standard because a failure to adequately consider these ancillary  co-pollutant
impacts could lead to an incorrect assessment of their net costs and benefits. Moreover, co-
pollutant impacts tend to accrue in the near term, while any effects from reduced climate
change mostly accrue over a time frame of several decades or longer.

       EPA typically quantifies and monetizes the health and environmental impacts related
to both PM and ozone in its  regulatory impact analyses (RIAs), when possible. However, we
were unable to do so in time for this proposal. EPA attempts to make emissions and air
quality modeling decisions early in the analytical process so that we can complete the
photochemical air quality modeling and use that data to inform the health and environmental
impacts analysis.  Resource  and time constraints precluded the Agency from completing this
work in time for the proposal.  Instead, EPA is using PM-related benefits-per-ton values as an
interim approach to estimating the PM-related benefits of the proposal. We also provide a
complete characterization of the health and environmental impacts that will be quantified and
monetized for the final ralemaking.

       This section is split into two sub-sections: the first presents the PM-related benefits-
per-ton values used to monetize the PM-related co-benefits associated with the proposal; the
second explains what PM- and ozone-related health and environmental impacts EPA will
quantify and monetize in the analysis for the final rale. EPA bases its analyses on peer-
reviewed studies of air quality and health and welfare effects and peer-reviewed studies of the
monetary values of public health and welfare improvements, and is generally consistent with

                                        7-31

-------
Draft Regulatory Impact Analysis
benefits analyses performed for the analysis of the final Ozone National Ambient Air Quality
Standard (NAAQS) and the final PM NAAQS analysis, as well as the recent Portland Cement
National Emissions Standards for Hazardous Air Pollutants (NESHAP) RIA (U.S. EPA,
2009a), and NO2 NAAQS (U.S.  EPA, 2009b).225'226'227'228

       Though EPA is characterizing the changes in emissions associated with toxic
pollutants, we will not be able to quantify or monetize the human health effects associated
with air toxic pollutants for either the proposal or the final rule analyses.  Please refer to
Chapter 5.5 for more information about the air toxics emissions impacts associated with the
proposed standards.

7.3.1 Economic Value of Reductions in Criteria Pollutants

        As described in Chapter 5.5, the proposed standards would reduce emissions of
several criteria and toxic pollutants and precursors.  In this analysis, EPA estimates the
economic value of the human health benefits associated with reducing PM2 5 exposure.  Due
to analytical limitations, this analysis does not estimate benefits related to other criteria
pollutants (such as ozone, NO2 or SO2) or toxic pollutants, nor does it monetize all of the
potential health and welfare effects associated with PM2 5.

       This analysis uses a "benefit-per-ton" method to estimate a selected suite of PM25-
related health benefits described below. These PM2 5 benefit-per-ton estimates provide the
total monetized human health benefits (the sum of premature mortality and premature
morbidity) of reducing one ton of directly emitted PM2 5,  or its precursors (such as NOx, SOx,
and VOCs), from a specified source. Ideally, the  human health benefits would be estimated
based on changes in ambient PM2 5 as determined by full-scale air quality modeling.
However, this modeling was not possible in the timeframe for this proposal.

       The dollar-per-ton estimates used in this analysis are provided in Table 7-4.  In the
summary of costs and benefits, Chapters.3 of this RIA, we present the monetized value of
PM-related improvements associated with the proposal.

    Table 7-4: Benefits-per-ton Values (2007S) Derived Using the ACS Cohort Study for PM-related
                 Premature Mortality (Pope et al., 2002)a and a 3% Discount Rate"
Year0
2015
2020
2030
2040
All Sources4
SOx
$28,000
$31,000
$36,000
$43,000
voc
$1,200
$1,300
$1,500
$1,800
Stationary (Non-EGU) Sources
NOx
$4,700
$5,100
$6,100
$7,200
Direct PM2.5
$220,000
$240,000
$280,000
$330,000
Mobile Sources
NOx
$4,900
$5,300
$6,400
$7,600
Direct PM2.5
$270,000
$290,000
$350,000
$420,000
a The benefit-per-ton estimates presented in this table are based on an estimate of premature mortality derived
from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six Cities study
(Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
b The benefit-per-ton estimates presented in this table assume a 3% discount rate in the valuation of premature
mortality to account for a twenty-year segmented cessation lag. If a 7% discount rate had been used, the values
would be approximately 9% lower.
                                          7-32

-------
                                                         Environmental and Health Impacts
 c Benefit-per-ton values were estimated for the years 2015, 2020, and 2030. For 2040, EPA extrapolated
 exponentially based on the growth between 2020 and 2030.
 d Note that the benefit-per-ton value for SOx is based on the value for Stationary (Non-EGU) sources; no SOx
 value was estimated for mobile sources.  The benefit-per-ton value for VOCs was estimated across all sources.

        The benefit per-ton technique has been used in previous analyses, including EPA's
 recent Ozone National Ambient Air Quality Standards (NAAQS) RIA (U.S. EPA, 2008a),229
 proposed Portland Cement National Emissions Standards for Hazardous Air Pollutants
 (NESHAP) RIA (U.S. EPA, 2009a),230 and proposed NO2 primary NAAQS (U.S. EPA,
 2009b).231   Table 7-5 shows the quantified and unquantified PM2 5-related co-benefits
 captured in those benefit-per-ton estimates.

	Table 7-5: Human Health and Welfare Effects of PM2.5	
   Pollutant /
     Effect
     Quantified and Monetized
       in Primary Estimates
          Unquantified Effects
         	Changes in:	
 PM2,
Adult premature mortality
Bronchitis: chronic and acute
Hospital admissions: respiratory and
cardiovascular
Emergency room visits for asthma
Nonfatal heart attacks (myocardial
infarction)
Lower and upper respiratory illness
Minor restricted-activity days
Work loss days
Asthma exacerbations (asthmatic
population)
Infant mortality
Subchronic bronchitis cases
Low birth weight
Pulmonary function
Chronic respiratory diseases other than chronic
bronchitis
Non-asthma respiratory emergency room visits
Visibility
Household soiling
        Consistent with the proposed NCh NAAQS,N the benefits estimates utilize the
 concentration-response functions as reported in the epidemiology literature.  Readers
 interested in reviewing the complete methodology for creating the benefit-per-ton estimates
 used in this analysis can consult the Technical Support Document (TSD)23 accompanying the
 recent final ozone NAAQS RIA (U.S. EPA, 2008a).  Readers can also refer to Farm et al.
 (2009)233 for a detailed description of the benefit-per-ton methodology.0 A more detailed
 description of the benefit-per-ton estimates is also provided in the TSD that accompanies this
 rulemaking.

        As described in the documentation for the benefit per-ton estimates cited above,
 national  per-ton estimates were developed for selected pollutant/source category
 NAlthough we summarize the main issues in this chapter, we encourage interested readers to see benefits chapter
 of the proposed primary NO2 NAAQS RIA for a more detailed description of recent changes to the PM benefits
 presentation and preference for the no-threshold model.
 0 The values included in this report are different from those presented in the article cited above. Benefits
 methods change to reflect new information and evaluation of the science. Since publication of the June 2009
 article, EPA has made two significant changes to its benefits methods: (1) We no longer assume that a threshold
 exists in PM-related models of health impacts; and (2) We have revised the Value of a Statistical Life to equal
 $6.3 million (year 2000$), up from an estimate of $5.5 million (year 2000$) used in the June 2009 report. Please
 refer to the following website for updates to the dollar-per-ton estimates:
 http://www.epa.gov/air/benmap/bpt.html
                                             7-33

-------
Draft Regulatory Impact Analysis

combinations.  The per-ton values calculated therefore apply only to tons reduced from those
specific pollutant/source combinations (e.g., NO2 emitted from mobile sources; direct PM
emitted from stationary sources).  Our estimate of PM2 5 benefits is therefore based on the
total direct PM2 5 and PM-related precursor emissions controlled by sector and multiplied by
each per-ton value.

       The benefit-per-ton coefficients in this analysis were derived using modified versions
of the health impact functions used in the PM NAAQS Regulatory Impact Analysis.
Specifically, this analysis uses the benefit-per-ton estimates first applied in the Portland
Cement NESHAP RIA (U.S. EPA, 2009a), which incorporated functions directly from  the
epidemiology studies without an adjustment for an assumed threshold. Removing the
threshold assumption is a key difference between the method used in this analysis to estimate
PM co-benefits and the methods used in analyses prior to EPA's proposed Portland Cement
NESHAP. The benefit-per-ton estimates now include incremental benefits down to the lowest
modeled PM2 5 air quality levels.

       PM-related mortality provides the majority (85-95%) of the monetized value in  each
benefit-per-ton estimate. As such, EPA deems it important to characterize the uncertainty
underlying the concentration-response (C-R) functions used in its benefits analyses of
regulations affecting PM levels. EPA has investigated methods to characterize uncertainty in
the relationship between PM25 exposure and premature mortality. EPA's final PM25 NAAQS
analysis provides a more complete picture about the overall uncertainty in PM2 5 benefits
estimates. For more information, please consult the PM2 5 NAAQS RIA (Table 5.5).
However, due to the limitations of the benefit-per-ton methodology employed here, the
quantitative uncertainty analysis related to the C-R relationship between PM2 5 and premature
mortality that EPA usually conducts in association with its benefits analysis was not
conducted for this proposal.

       Typically, the premature mortality-related effect coefficients that underlie the benefits-
per-ton estimates are drawn from epidemiology studies that examine two large population
cohorts: the American Cancer Society cohort (Pope et al., 2002)234 and the Harvard Six Cities
cohort (Laden et al., 2006).235 The concentration-response (C-R) function developed from the
extended analysis of American Cancer Society (ACS) cohort, as reported in Pope et al.
(2002), has previously been used by EPA to generate its primary benefits estimate. The
extended analysis of the Harvard Six Cities cohort, as reported by Laden et al (2006), was
published after the  completion of the Staff Paper for the 2006 PM2 5 NAAQS and has been
used as an alternative estimate in the PM2 5 NAAQS RIA and PM2 5 co-benefits estimates in
analyses completed since the PM2 5 NAAQS. These are logical choices for anchor points
when presenting PM-related benefits because, although both studies are well designed and
peer reviewed, there are strengths and weaknesses inherent in each, which argues for using
both studies to generate benefits estimates.  Using the alternate relationships between PM2 5
and premature mortality supplied by experts as part of EPA's 206 Expert Elicitation Study,
higher and lower benefits estimates are plausible, but most of the expert-based estimates fall
between the two epidemiology-based estimates (Roman et al., 2008; lEc, 2006).236'237
However, due to the analytical limitations associated with this analysis, we have chosen to use
the benefit-per-ton value derived from the ACS study and note that benefits would be
                                        7-34

-------
                                                    Environmental and Health Impacts

approximately 145% (or nearly two-and-a-half times) larger if the Harvard Six Cities values
were used.

       As a note to those who might be comparing the benefits estimates in this rule to those
in previous EPA analyses, it is the nature of benefits analyses for assumptions and methods to
evolve over time to reflect the most current interpretation of the scientific and economic
literature. For a period of time (2004-2008), EPA's Office of Air and Radiation (OAR)
valued mortality risk reductions using a value of statistical life (VSL) estimate derived from a
limited analysis of some of the available studies. OAR arrived at a VSL using a range of $1
million to $10 million (2000$) consistent with two meta-analyses of the wage-risk literature.
The $1 million value represented the lower end of the interquartile range from the Mrozek and
Taylor (2002)238 meta-analysis of 33 studies.  The $10 million value represented the upper
end of the interquartile range from the Viscusi and Aldy (2003)239 meta-analysis of 43 studies.
The mean estimate of $5.5 million (2000$) was also consistent with the mean VSL of $5.4
million estimated in the Kochi et al. (2006)240 meta-analysis. However, the Agency neither
changed its official guidance on the use of VSL in rule-makings nor subjected the interim
estimate to a scientific peer-review process through the Science Advisory Board (SAB) or
other peer-review group.

       Until updated guidance is available, EPA determined that a single, peer-reviewed
estimate applied consistently best reflects the SAB-EEAC advice it has received.  Therefore,
EPA has decided to apply the VSL that was vetted and endorsed by the SAB in the Guidelines
for Preparing Economic Analyses (U.S. EPA, 2000)241 while they continue efforts to update
their guidance on this issue.p This approach calculates a mean value across VSL estimates
derived from 26 labor market and contingent valuation studies published between 1974 and
1991. The mean VSL across these studies is $6.3 million (2000$). The dollar-per-ton
estimates used in this analysis are based on this VSL.

       The benefit-per-ton estimates are subject to a number of assumptions and
uncertainties.

           •   They do not reflect local variability in population density, meteorology,
              exposure, baseline health incidence rates, or other local factors that might lead
              to an overestimate  or underestimate of the actual benefits of controlling fine
              particulates. EPA will conduct full-scale air quality modeling for the final
              rulemaking in an effort to capture this variability.  Please refer to Section VILE
              for a description of EPA's modeling plans and to Section VIII.G.2  for the
              description of the quantification and monetization of health impacts for the
              FRM.
           •   This analysis assumes that all fine particles, regardless of their chemical
              composition,  are equally potent in causing premature mortality. This is an
p In the (draft) update of the Economic Guidelines (U.S. EPA, 2008c), EPA retained the VSL endorsed by the
SAB with the understanding that further updates to the mortality risk valuation guidance would be forthcoming
in the near future. Therefore, this report does not represent final agency policy. The draft update of the
Economic Guidelines is available on the Internet at .

                                          7-35

-------
Draft Regulatory Impact Analysis
              important assumption, because PM2 5 produced via transported precursors
              emitted from stationary sources may differ significantly from direct PM2 5
              released from diesel engines and other industrial sources, but no clear scientific
              grounds exist for supporting differential effects estimates by particle type.
              This analysis assumes that the health impact function for fine particles is linear
              within the range of ambient concentrations under consideration. Thus, the
              estimates include health benefits from reducing fine particles in areas with
              varied concentrations of PM2 5, including both regions that are in attainment
              with fine particle standard and those that do not meet the standard down to the
              lowest modeled concentrations.
              There are several health benefits categories that we were unable to quantify
              due to limitations associated with using benefits-per-ton estimates, several of
              which could be substantial.  Because the NOX and VOC emission reductions
              associated with this proposal are also precursors to ozone, reductions in NOx
              and VOC would also reduce ozone formation and the health effects associated
              with ozone exposure.  Unfortunately, benefits-per-ton estimates do not exist
              due to issues associated with the complexity of the atmospheric air chemistry
              and nonlinearities associated with ozone formation.  The PM-related benefits-
              per-ton estimates also do not include any human welfare or ecological benefits.
              Please refer to  VIII. G.2 for a description of the quantification and monetization
              of health impact for the FRM and a description of the unquantified co-pollutant
              benefits associated with this rulemaking.
              There are many uncertainties associated with the health impact functions used
              in this modeling effort. These include: within-study variability (the precision
              with which a given study estimates the relationship between air quality
              changes and health effects); across-study variation (different published studies
              of the same pollutant/health effect relationship typically do not report identical
              findings and in some instances the differences are substantial); the application
              of C-R functions nationwide  (does not account for any relationship between
              region and health effect, to the extent that such a relationship exists);
              extrapolation of impact functions across population (we assumed that certain
              health impact functions applied to age ranges broader than that considered in
              the original epidemiological study); and various uncertainties in the C-R
              function, including causality and thresholds. These uncertainties may under-
              or over-estimate benefits.
              EPA has investigated methods to characterize uncertainty in the relationship
              between PM2 5 exposure and premature mortality. EPA's final PM25 NAAQS
              analysis provides a more complete picture about the overall uncertainty in
              PM2 5 benefits  estimates. For more information, please consult the PM2 5
              NAAQS RIA (Table 5.5).
              The benefit-per-ton estimates used in this analysis incorporate projections of
              key variables, including atmospheric conditions, source level emissions,
              population, health baselines and incomes, technology. These projections
              introduce some uncertainties  to the benefit per ton estimates.
                                         7-36

-------
                                                   Environmental and Health Impacts

             As described above, using the benefit-per-ton value derived from the ACS
             study (Pope et al, 2002) alone provides an incomplete characterization of
             PM2 5 benefits. When placed in the context of the Expert Elicitation results,
             this estimate falls toward the lower end of the distribution. By contrast, the
             estimated PM2 5 benefits using the coefficient reported by Laden in that
             author's reanalysis of the Harvard Six Cities cohort fall toward the upper end
             of the Expert Elicitation distribution results.
       As mentioned above, emissions changes and benefits-per-ton estimates alone are not a
good indication of local or regional air quality and health impacts, as there may be localized
impacts associated with the proposed rulemaking.  Additionally, the atmospheric chemistry
related to ambient concentrations of PM2 5, ozone and air toxics is very complex.  Full-scale
photochemical modeling is therefore necessary to provide the needed spatial and temporal
detail to more completely and accurately estimate the changes in ambient levels of these
pollutants and their associated health and welfare impacts. As discussed above, timing and
resource constraints precluded EPA from conducting a full-scale photochemical air quality
modeling analysis in time for the NPRM. For the final rule, however, a national-scale air
quality modeling analysis will be performed to analyze the impacts of the standards  on PM25,
ozone,  and selected air toxics. The benefits analysis plan for the final rulemaking is discussed
in the next section.

7.3.2  Human Health and Environmental Benefits for the Final Rule

7.3.2.1  Human Health and Environmental Impacts

       To model the ozone and PM air quality benefits of the final rule, EPA will use the
Community Multiscale Air Quality (CMAQ) model (see Section 7.2.3.1 for a description of
the CMAQ model). The modeled ambient air quality data will serve as an input to the
Environmental Benefits Mapping and  Analysis Program (BenMAP).242 BenMAP is a
computer program developed by EPA that integrates a number of the modeling elements used
in previous RIAs (e.g., interpolation functions, population projections, health impact
functions, valuation functions, analysis and pooling methods) to translate modeled air
concentration estimates into health effects incidence estimates and monetized benefits
estimates.
       Table 7-6 lists the co-pollutant health effect exposure-response functions we will use
to quantify the co-pollutant incidence impacts associated with the final light-duty vehicles
standard.
                                        7-37

-------
Draft Regulatory Impact Analysis
Table 7-6: Health Impact Functions Used in BenMAP to Estimate Impacts of PM2.5 and Ozone Reductions
ENDPOINT
POLLUTANT
STUDY
STUDY POPULATION
Premature Mortality
Premature mortality —
daily time series
Premature mortality
— cohort study, all-
cause
Premature mortality,
total exposures
Premature mortality
— all-cause
03
PM2.5
PM2.5
PM2.5
Multi-city
Bell et al (2004) (NMMAPS study)243 - Non-
accidental
Huang et al (2005)244 - Cardiopulmonary
Schwartz (2005 )245 -Non-accidental
Meta-analvses:
Bell et al (2005 )246 - All cause
Ito et al (2005)247 -Non-accidental
Levy et al (2005)248 - All cause
Popeetal. (2002)249
Laden et al. (2006)250
Expert Elicitation (lEc, 2006)251
Woodruff etal. (1997)252
All ages
>29 years
>25 years
>24 years
Infant (<1 year)
Chronic Illness
Chronic bronchitis
Nonfatal heart attacks
PM2.5
PM2.5
Abbey et al. (1995 )253
Peters etal. (200 1)254
>26 years
Adults (> 18 years)
Hospital Admissions
RespirFSatory
Cardiovascular
03
PM25

PM2.5
PM2.5
PM2.5
PM2.5
PM2.5
Pooled estimate:
Schwartz (1995) - ICD 460-519 (all resp)255
Schwartz (1994a; 1994b) - ICD 480-486
(pneumonia)256'257
Moolgavkar et al. (1997) - ICD 480-487
(pneumonia)258
Schwartz (1994b) - ICD 491-492, 494-496
(COPD)
Moolgavkar et al. (1997) - ICD 490-496
(COPD)
Burnett etal. (200 1)259
Pooled estimate:
Moolgavkar (2003)— ICD 490-496 (COPD)260
Ito (2003)— ICD 490-496 (COPD)261
Moolgavkar (2000)— ICD 490-496 (COPD)262
Ito (2003)— ICD 480-486 (pneumonia)
Sheppard (2003)— ICD 493 (asthma)263
Pooled estimate:
Moolgavkar (2003)— ICD 390-429 (all
cardiovascular)
Ito (2003)— ICD 410-414, 427-428 (ischemic
heart disease, dysrhythmia, heart failure)
Moolgavkar (2000)— ICD 390-429 (all
cardiovascular)
>64 years
<2 years
>64 years
20-64 years
>64 years
<65 years
>64 years
20—64 years
                                          7-38

-------
                                                         Environmental and Health Impacts
Asthma-related ER
visits
Asthma-related ER
visits (con't)
03
PM2.5
Pooled estimate:
Jaffeetal(2003)264
Peel etal (2005 )265
Wilson et al (2005)266
Nonisetal. (1999)267
5-34 years
All ages
All ages
0-1 8 years
Other Health Endpoints
Acute bronchitis
Upper respiratory
symptoms
Lower respiratory
symptoms
Asthma exacerbations
Work loss days
School absence days
Minor Restricted
Activity Days
(MRADs)
PM2.5
PM2.5
PM2.5
PM2.5
PM2.5
03
03
PM2.5
Dockeryetal. (1996)268
Pope etal. (199 1)269
Schwartz andNeas (2000)270
Pooled estimate:
Ostro et al. (200 1)271 (cough, wheeze and
shortness of breath)
Vedal et al. (1998) (cough)
Ostro (1987)273
Pooled estimate:
Gillilandetal. (200 1)274
Chen etal. (2000)275
Ostro and Rothschild (1989)276
Ostro and Rothschild (1989)
8—12 years
Asthmatics, 9-1 1
years
7-14 years
6-1 8 years3
18-65 years
5-17 yearsb
18-65 years
18-65 years
Notes:
a The original study populations were 8 to 13 for the Ostro et al. (2001) study and 6 to 13 for the Vedal et al.
(1998) study.  Based on advice from the Science Advisory Board Health Effects Subcommittee (SAB-HES), we
extended the applied population to 6 to 18, reflecting the common biological basis for the effect in children in
the broader age group. See: U.S. Science Advisory Board. 2004. Advisory Plans for Health Effects Analysis in
the Analytical Plan for EPA's Second Prospective Analysis -Benefits and Costs of the Clean Air Act, 1990—
2020. EPA-SAB-COUNCIL-ADV-04-004. See also National Research Council (NRC).  2002. Estimating the
Public Health Benefits of Proposed Air Pollution Regulations. Washington, DC: The National Academies
Press.
b Gillilandetal. (2001) studied children aged 9 and 10. Chen et al. (2000) studied children 6 to 11.  Based on
recent advice from the National Research Council and the EPA SAB-HES, we have calculated reductions in
school absences for all school-aged children based on the biological similarity between children aged 5 to 17.

7.3.2.2   Monetized Estimates of Impacts of Reductions in Co-Pollutants

        Table 7-7 presents the monetary values  we will apply to changes in the incidence of
health and welfare effects associated with reductions in non-GHG pollutants that will  occur
when these GHG control strategies are finalized.
                                             7-39

-------
Draft Regulatory Impact Analysis
     	Table 7-7: Valuation Metrics Used in BenMAP to Estimate Monetary Co-Benefits
                                                                          Valuation
     Endpoint	Valuation Method	(2000S)
     Premature mortality
Assumed Mean VSL
$6,300,000
     Chronic Illness
      Chronic Bronchitis
      Myocardial Infarctions, Nonfatal
WTP: Average Severity
Medical Costs Over 5 Years. Varies by
age and discount rate. Russell
(1998)277
Medical Costs Over 5 Years. Varies by
age and discount rate. Wittels
(1990)278	
 $340,482
     Hospital Admissions
       Respiratory, Age 65+
       Respiratory, Ages 0-2
COI: Medical Costs + Wage Lost
COI: Medical Costs
 $18,353
  $7,741
       Chronic Lung Disease (less
       Asthma)
COI: Medical Costs + Wage Lost
 $12,378
Pneumonia
Asthma
Cardiovascular
ER Visits, Asthma
Other Health Endpoints
Acute Bronchitis
Upper Respiratory Symptoms
Lower Respiratory Symptoms
Asthma Exacerbation
COI: Medical Costs + Wage Lost
COI: Medical Costs + Wage Lost
COI: Medical Costs + Wage Lost (20-
64)
COI: Medical Costs + Wage Lost (65-
99)
COI: Smith etal.(1997)279
COI: Standfordetal. (1999)280
WTP: 6 Day Illness, CV Studies
WTP: 1 Day, CV Studies
WTP: 1 Day, CV Studies
WTP: Bad Asthma Day, Rowe and
$14,693
$6,634
$22,778
$21,191
$312
$261
$356
$25
$16
$43
      Work Loss Days
      Minor Restricted Activity Days
      School Absence Days
      Worker Productivity
              OO 1
Chestnut (1986)
Median Daily Wage, County-Specific
WTP: 1 Day, CV Studies
Median Daily Wage, Women 25+
Median Daily Wage, Outdoor
Workers, County-Specific	
   $51
   $75
     Environmental Endpoints
       Recreational Visibility
WTP: 86 Class I Areas
      Source: Dollar amounts for each valuation method were extracted from BenMAP version 3.0.

7.3.2.3   Other Unquantified Health and Environmental Impacts

         In addition to the co-pollutant health and environmental impacts we will quantify for
the analysis of the Light-Duty Vehicle GHG standard, there are a number of other health and
human welfare endpoints that we will not be able to quantify because of current limitations in
the methods or available data. These impacts are associated with emissions of air toxics
(including benzene, 1,3-butadiene, formaldehyde, acetaldehyde, acrolein, and  ethanol),
ambient ozone, and ambient PM2 5 exposures. For example, 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 (i.e.,
                                           7-40

-------
                                                    Environmental and Health Impacts
changes in heart rate variability).  In addition, we are currently unable to quantify a number
of known welfare effects, including reduced acid and paniculate deposition damage to
cultural monuments and other materials, and environmental benefits due to reductions of
impacts of eutrophication in coastal areas.  For air toxics, the available tools and methods to
assess risk from mobile sources at the national scale are not adequate for extrapolation to
benefits assessment.  In addition to inherent limitations in the tools for national-scale
modeling of air toxics and exposure, there is a lack of epidemiology data for air toxics in the
general population.  Table 7-8 lists these unqualified health and environmental impacts.
                  Table 7-8: Unquantified and Non-Monetized Potential Effects
POLLUTANT/EFFECTS
Ozone Health
Ozone Welfare
PM Healthb
PM Welfare
Nitrogen and Sulfate
Deposition Welfare
CO Health
Hydrocarbon (HC)/Toxics
Health"
EFFECTS NOT INCLUDED IN ANALYSIS - CHANGES IN:
Chronic respiratory damage
Premature aging of the lungs
Non-asthma respiratory emergency room visits
Exposure to UVb (+/-)d
Yields for
-commercial forests
-some fruits and vegetables
-non-commercial crops
Damage to urban ornamental plants
Impacts on recreational demand from damaged forest aesthetics
Ecosystem functions
Exposure to UVb (+/-)
Premature mortality - short term exposures
Low birth weight
Pulmonary function
Chronic respiratory diseases other than chronic bronchitis
Non-asthma respiratory emergency room visits
Exposure to UVb (+/-)
Residential and recreational visibility in non-Class I areas
Soiling and materials damage
Damage to ecosystem functions
Exposure to UVb (+/-)
Commercial forests due to acidic sulfate and nitrate deposition
Commercial freshwater fishing due to acidic deposition
Recreation in terrestrial ecosystems due to acidic deposition
Existence values for currently healthy ecosystems
Commercial fishing, agriculture, and forests due to nitrogen deposition
Recreation in estuarine ecosystems due to nitrogen deposition
Ecosystem functions
Passive fertilization
Behavioral effects
Cancer (benzene, 1,3-butadiene, formaldehyde, acetaldehyde, ethanol)
Anemia (benzene)
Disruption of production of blood components (benzene)
Reduction in the number of blood platelets (benzene)
Excessive bone marrow formation (benzene)
Depression of lymphocyte counts (benzene)
Reproductive and developmental effects (1,3-butadiene, ethanol)
Irritation of eyes and mucus membranes (formaldehyde)
Respiratory irritation (formaldehyde)
Asthma attacks in asthmatics (formaldehyde)
                                         7-41

-------
Draft Regulatory Impact Analysis
                        Asthma-like symptoms in non-asthmatics (formaldehyde)
                        Irritation of the eyes, skin, and respiratory tract (acetaldehyde)
                        Upper respiratory tract irritation and congestion (acrolein)
HC/Toxics Welfare
Direct toxic effects to animals
Bioaccumulation in the food chain
Damage to ecosystem function
Odor
  In addition to primary economic endpoints, there are a number of biological responses that have been
associated with ozone health effects including increased airway responsiveness to stimuli, inflammation in the
lung, acute inflammation and respiratory cell damage, and increased susceptibility to respiratory infection. The
public health impact of these biological responses may be partly represented by our quantified endpoints.
 In addition to primary economic endpoints, there are a number of biological responses that have been
associated with PM health effects including morphological changes and altered host defense mechanisms. The
public health impact of these biological responses may be partly represented by our quantified endpoints.
 While some of the effects of short-term exposures are likely to be captured in the estimates, there may be
premature mortality due to short-term exposure to PM not captured in the cohort studies used in this analysis.
However, the PM mortality results derived from the expert elicitation do take  into account premature mortality
effects of short term exposures.
 May result in benefits or disbenefits.
 Many of the key hydrocarbons related to this rule are also hazardous air pollutants listed in the Clean Air Act.
Please refer to Chapter 8.4 for additional information on the health effects of air toxics.
 Please refer to Chapter 8.4 for additional information on the welfare effects of air toxics.

        In addition to the co-pollutant health and environmental impacts we will quantify for
the analysis of the final standard, there are a number of other health and human welfare
endpoints that we will not be able to quantify or monetize because of current limitations in the
methods or available data. These impacts are associated with emissions of air toxics
(including benzene, 1,3-butadiene, formaldehyde, acetaldehyde, acrolein, andethanol),
ambient ozone, and ambient PM2 5 exposures. Chapter 7.3 of the RIA lists these unquantified
health and environmental impacts.

        While there will be impacts associated with air toxic pollutant emission changes that
result from the final standard, we will not attempt to monetize those impacts. This is
primarily because currently available tools and methods to assess air toxics risk from mobile
sources at the national scale are not adequate for extrapolation to incidence estimations or
benefits assessment. The best suite of tools and methods currently available for assessment at
the national scale are those used in the National-Scale Air Toxics Assessment (NATA). The
EPA Science Advisory Board specifically commented in their review of the 1996 NATA that
these tools were not yet ready for use in a national-scale benefits analysis, because they did
not consider the full distribution of exposure and risk, or address sub-chronic health effects.282
While EPA has since improved the tools, there remain critical limitations for estimating
incidence and  assessing benefits of reducing mobile source air toxics. EPA continues to work
to address these limitations; however, we do not anticipate having methods and tools available
                                             7-42

-------
                                                   Environmental and Health Impacts

for national-scale application in time for the analysis of the final rules.Q
7.4  Changes in Global Mean Temperature and Sea-Level Rise Associated
    with the Proposal's GHG Emissions Reductions

7.4.1 Introduction

        Based on modeling analysis performed by the EPA, reductions in CO2 and other
GHGs associated with the Proposal will affect climate change projections. Because GHGs
mix well in the atmosphere and have long atmospheric lifetimes, changes in GHG emissions
will affect atmospheric concentrations of greenhouse gases and future climate for decades to
centuries. Two common indicators of climate change are global mean surface temperature and
sea-level rise. This section estimates the response in global mean surface temperature and sea-
level rise projections to the estimated net global GHG emissions reductions associated with
the Proposal (see Chapter 5 for the estimated net reductions in global emissions over time by
GHG).

7.4.2 Estimated Projected Reductions in Global Mean Surface Temperature and
      Sea-Level Rise

        We estimated changes in projected atmospheric  CCh concentrations, global mean
surface temperature and sea-level rise to 2100 using the MiniCAM (Mini Climate Assessment
Model) integrated assessment model283 coupled with the MAGICC (Model for the
Assessment of Greenhouse-gas Induced Climate Change) simple climate model.284 MiniCAM
was used to create the globally and temporally consistent set of climate relevant variables
required for running MAGICC. MAGICC was then used to estimate the change in the
atmospheric CCh concentrations, global mean surface temperature and sea-level rise over
time. Given the magnitude of the estimated emissions reductions associated with the  rule, a
simple climate model such as MAGICC is reasonable for estimating the atmospheric and
climate response.

      An emissions scenario for the proposal was developed by applying the proposal's
estimated emissions reductions to the MiniCAM reference (no climate policy or baseline)
                                                                         9SS
scenario (used as the basis for the Representative Concentration Pathway RCP4.5  ).
Specifically, the CCh, N2O, CH4, and HFC-134a emissions reductions from Chapter 5 were
       Q In April, 2009, EPA hosted a workshop on estimating the benefits or reducing hazardous air
pollutants. This workshop built upon the work accomplished in the June 2000 Science Advisory Board/EPA
Workshop on the Benefits of Reductions in Exposure to Hazardous Air Pollutants, which generated thoughtful
discussion on approaches to estimating human health benefits from reductions in air toxics exposure, but no
consensus was reached on methods that could be implemented in the near term for a broad selection of air toxics.
Please visit http://epa.gov/air/toxicair/2009workshop.html for more information about the workshop and its
associated materials.
                                        7-43

-------
Draft Regulatory Impact Analysis
applied as net reductions to the MiniCAM global baseline net emissions for each GHG. All
emissions reductions were assumed to begin in 2012, with zero emissions change in 201 1 and
linearly increasing to equal the value supplied (in Chapter 5) for 2020. The emissions
reductions past 2050 were scaled with total U.S. road transportation fuel consumption from
the MiniCAM reference scenario. Using MAGICC, the atmospheric CO2 concentration, the
global mean temperature, and sea-level change were projected at five-year time steps to 2 100
for both the reference (no climate policy) scenario and the emissions scenario specific to the
Proposal.  To capture some of the uncertainty in the climate system, the changes in projected
temperatures and sea level were estimated across the most current Intergovernmental Panel on
Climate Change (IPCC) range of climate sensitivities, 1.5°C to 6.0°C.286

       To compute the reductions in atmospheric CCh concentration, temperature, and sea-
level rise specifically attributable to the Proposal, the output from the Proposal's emissions
scenario was subtracted from the reference (no policy or baseline) emissions case scenario.
As a result of the Proposal's emissions reductions, the atmospheric CO2 concentration is
projected to be reduced by approximately 2.9 to 3.2 parts per million (ppm), the global mean
temperature is projected to be reduced by approximately 0. 007-0. 016°C by 2100 and global
mean sea-level rise is projected to be reduced by approximately 0.06-0.15 cm by 2100.

       Figure 7-4 provides the estimated reductions in projected global mean surface
temperatures associated with the Proposal. Figure 7-5 provides the estimated reductions in
global mean sea-level rise associated with the Proposal.
                  Global Mean Temperature Change
                    2000
              0.0000  _
2050
2100
             -0.0050 -
          a -0.0100 -

          •S -0.0150 -
             -0.0200
                       ClimSens=1.5
                      • ClimSens=2.0
                       ClimSens=2.5
                       ClimSens=3.0
                      •ClimSens=4.5
                       ClimSens=6.0
                                Year
 Figure 7-4 Estimated Projected Reductions in Global Mean Surface Temperatures from Baseline for the
           Proposed Vehicles Rulemaking (for climate sensitivities ranging from 1.5-6°C)
                                        7-44

-------
                                                   Environmental and Health Impacts
                Global Mean Sea Level Rise Change
                              2050
2100
              -0.20
                                                      ClimSens=1.5
                                                     • ClimSens=2.0
                                                      ClimSens=2.5
                                                      ClimSens=3.0
                                                     •ClimSens=4.5
                                                      ClimSens=6.0
                              Year
    Figure 7-5 Estimated Projected Reductions in Global Mean Sea-Level Rise from Baseline for the
           Proposed Vehicles Rulemaking (for climate sensitivities ranging from 1.5-6°C)

       The results in both Figure 7-4 and Figure 7-5 show a relatively small reduction in the
projected global mean surface temperature and sea level respectively, across all climate
sensitivities. The projected reductions are small relative to the IPCC's 2100 "best estimates"
for global mean temperature increases (1.8 - 4.0°C) and sea-level rise (0.20-0.59m) for all
global GHG emissions sources for a range of emissions scenarios.287   These projected
reductions are proportionally representative of changes to U.S. GHG emissions in the
transportation sector. While not formally estimated for the proposed rulemaking, a  reduction
in projected global mean temperature and sea-level rise implies a reduction in the risks
associated with of climate change. Both figures illustrate that the distribution for projected
global mean temperature and sea-level rise increases has shifted down. The benefits of GHG
emissions reductions can be characterized both qualitatively and quantitatively, some of
which can be monetized (seeChapter 7.5). There are substantial uncertainties in modeling the
global risks of climate change, which complicates quantification and benefit-cost assessments.
Changes in climate variables are a meaningful proxy for changes in the risk of all potential
impacts-including those that can be monetized, and those that have not been monetized but
can be quantified in physical terms (e.g., water availability), as well as those that have not yet
been quantified or are  extremely difficult to quantify (e.g., respectively forest disturbance and
catastrophic events such as collapse of large ice sheets and subsequent sea-level rise).
                                         7-45

-------
Draft Regulatory Impact Analysis

7.5  SCC and GHG Benefits

       We assigned a monetary value to reductions in CCh 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, however, 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. Furthermore, some of the likely and potential damages from climate change—for
example, the loss of endangered species—are generally not included in current SCC
estimates. These omissions may turn out to be significant, in the sense that they may mean
that the best current estimates are too low.  As noted by the IPCC Fourth Assessment Report,
"It is very likely that globally aggregated figures underestimate the damage costs because they
cannot include many non-quantifiable impacts."288

       Today's joint proposals present a set of interim SCC values reflecting a federal
interagency group's interpretation of the relevant climate economics literature. The interim
SCC values, which reflect an interim interpretation of the current literature, are derived using
several discount rates. The interim SCC values include:

          •   $5 (based on a 5% discount rate);

          •   $ 10(5% using Newell-Pizer adjustment);,

          •   $20(average SCC value from the average SCC estimates based on 5% and
              3%);

          •   $34  (3%);

          •   $56  (3% using Newell-Pizer adjustment).

       These interim SCC values are in 2007 dollars, and are based on a CCh emissions
change of 1 metric ton in 2007. Section III.H.6 of the Preamble provides a complete
discussion about SCC and the interim set of values.

       The tables below summarize the total GHG benefits for the lifetime of the rule, which
are calculated by using the five interim SCC values. Specifically, total monetized benefits in
each year are calculated by multiplying the marginal benefits estimates per metric ton of CCh
(the SCC) by the reductions in CCh for that year.  We have also approximated the total
monetized benefits for non-CCh GHGs by multiplying the SCC value by the  reductions in
non-CCh GHGs for that year. Marginal benefit estimates per metric ton of non-CO2 GHGs
are currently unavailable, but work is on-going to monetize benefits related to the mitigation
of other non-CO2 GHGs.
                                        7-46

-------
                                                        Environmental and Health Impacts
Table 7-9: Upstream and Downstream CO2 Benefits for the Given SCC Value, Calendar Year Analysis (Millions of
                                          2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
5%
$35
$89
$168
$281
$439
$601
$768
$939
$1,112
$1,288
$1,465
$1,645
$1,827
$2,014
$2,200
$2,386
$2,568
$2,749
$2,931
$3,117
$3,306
$3,501
$3,702
$3,912
$4,129
$4,357
$4,594
$4,844
$5,104
$5,379
$5,668
$5,973
$6,294
$6,632
$6,987
$7,362
$7,756
$8,171
$8,607
$62,100
$25,600
5%
NEWELL-
PIZER
$70
$179
$336
$562
$878
$1,203
$1,536
$1,878
$2,225
$2,577
$2,929
$3,290
$3,654
$4,027
$4,400
$4,772
$5,136
$5,497
$5,861
$6,234
$6,611
$7,002
$7,404
$7,824
$8,258
$8,715
$9,188
$9,687
$10,209
$10,758
$11,336
$11,945
$12,587
$13,263
$13,975
$14,724
$15,512
$16,342
$17,214
$124,200
$51,200
FROM 3%
AND 5%
$132
$340
$639
$1,068
$1,668
$2,285
$2,919
$3,568
$4,227
$4,896
$5,566
$6,250
$6,942
$7,652
$8,359
$9,067
$9,759
$10,445
$11,137
$11,844
$12,562
$13,305
$14,068
$14,865
$15,691
$16,558
$17,458
$18,406
$19,396
$20,441
$21,539
$22,696
$23,916
$25,200
$26,552
$27,976
$29,473
$31,049
$32,706
$236,000
$97,200
3%
$230
$590
$1,109
$1,855
$2,898
$3,968
$5,069
$6,197
$7,342
$8,503
$9,666
$10,855
$12,057
$13,290
$14,519
$15,748
$16,949
$18,141
$19,342
$20,571
$21,818
$23,108
$24,434
$25,819
$27,253
$28,758
$30,322
$31,969
$33,688
$35,502
$37,409
$39,420
$41,538
$43,768
$46,117
$48,590
$51,191
$53,927
$56,805
$409,800
$168,800
3%
NEWELL-
PIZER
$383
$984
$1,849
$3,092
$4,829
$6,614
$8,449
$10,328
$12,237
$14,172
$16,111
$18,092
$20,095
$22,151
$24,198
$26,247
$28,249
$30,235
$32,237
$34,285
$36,363
$38,513
$40,723
$43,031
$45,421
$47,930
$50,536
$53,281
$56,147
$59,170
$62,349
$65,700
$69,229
$72,947
$76,861
$80,983
$85,318
$89,878
$94,675
$683,100
$281,300
                                             7-47

-------
    Draft Regulatory Impact Analysis
Table 7-10: Upstream and Downstream non-CO2 GHG Benefits for the Given SCC Value, Calendar Year Analysis
                                     (Millions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
5%
$5
$13
$24
$38
$55
$72
$91
$110
$133
$154
$175
$198
$220
$243
$267
$290
$313
$334
$358
$379
$400
$422
$443
$465
$487
$510
$533
$557
$581
$607
$633
$661
$689
$719
$749
$781
$814
$847
$883
$7,100
$3,000
5%
NEWELL-
PIZER
$11
$27
$48
$76
$109
$144
$181
$220
$266
$308
$351
$395
$440
$487
$533
$580
$626
$668
$716
$758
$801
$843
$886
$930
$974
$1,019
$1,066
$1,114
$1,163
$1,214
$1,267
$1,321
$1,378
$1,437
$1,498
$1,561
$1,627
$1,695
$1,765
$14,200
$6,000
FROM 3%
AND 5%
$21
$51
$91
$145
$208
$274
$344
$417
$506
$585
$667
$751
$837
$925
$1,013
$1,102
$1,189
$1,269
$1,361
$1,441
$1,521
$1,602
$1,683
$1,766
$1,851
$1,937
$2,025
$2,116
$2,209
$2,306
$2,407
$2,511
$2,619
$2,730
$2,846
$2,967
$3,091
$3,220
$3,354
$27,100
$11,400
3%
$36
$88
$158
$252
$361
$476
$597
$725
$879
$1,016
$1,158
$1,304
$1,453
$1,606
$1,760
$1,913
$2,064
$2,204
$2,364
$2,503
$2,642
$2,782
$2,924
$3,067
$3,214
$3,364
$3,517
$3,675
$3,836
$4,005
$4,180
$4,361
$4,548
$4,742
$4,944
$5,153
$5,369
$5,593
$5,826
$47,000
$19,700
3%
NEWELL-
PIZER
$60
$147
$263
$419
$601
$793
$996
$1,208
$1,465
$1,694
$1,929
$2,173
$2,422
$2,677
$2,934
$3,189
$3,441
$3,674
$3,940
$4,172
$4,403
$4,637
$4,873
$5,112
$5,357
$5,607
$5,862
$6,124
$6,394
$6,676
$6,967
$7,268
$7,580
$7,904
$8,240
$8,588
$8,949
$9,322
$9,710
$78,400
$32,900
                                             7-48

-------
                                                        Environmental and Health Impacts
Table 7-11: Upstream and Downstream CO2-Equivalent Benefits for the Given SCC Value, Calendar Year Analysis
                                     (Millions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
5%
$40
$103
$192
$319
$494
$673
$859
$1,049
$1,246
$1,442
$1,640
$1,842
$2,047
$2,257
$2,466
$2,676
$2,881
$3,083
$3,289
$3,496
$3,706
$3,923
$4,145
$4,377
$4,616
$4,867
$5,127
$5,400
$5,686
$5,986
$6,301
$6,633
$6,983
$7,350
$7,736
$8,143
$8,570
$9,018
$9,490
$69,200
$28,600
5%
NEWELL-
PIZER
$81
$206
$384
$638
$987
$1,347
$1,717
$2,098
$2,491
$2,885
$3,280
$3,685
$4,094
$4,514
$4,933
$5,352
$5,762
$6,165
$6,578
$6,992
$7,412
$7,845
$8,290
$8,753
$9,232
$9,734
$10,254
$10,801
$11,371
$11,972
$12,603
$13,267
$13,965
$14,700
$15,473
$16,286
$17,139
$18,036
$18,979
$138,400
$57,100
FROM 3%
AND 5%
$153
$391
$729
$1,213
$1,876
$2,559
$3,263
$3,985
$4,734
$5,481
$6,232
$7,001
$7,779
$8,577
$9,373
$10,169
$10,947
$11,714
$12,498
$13,285
$14,083
$14,906
$15,751
$16,631
$17,541
$18,494
$19,483
$20,522
$21,605
$22,747
$23,945
$25,207
$26,534
$27,930
$29,398
$30,943
$32,565
$34,269
$36,060
$263,000
$108,500
3%
$266
$678
$1,267
$2,107
$3,258
$4,444
$5,667
$6,922
$8,222
$9,520
$10,824
$12,159
$13,510
$14,897
$16,279
$17,661
$19,014
$20,345
$21,707
$23,074
$24,460
$25,890
$27,357
$28,886
$30,467
$32,122
$33,839
$35,643
$37,525
$39,507
$41,589
$43,781
$46,086
$48,510
$51,060
$53,743
$56,560
$59,520
$62,631
$456,900
$188,500
3%
NEWELL-
PIZER
$443
$1,131
$2,112
$3,511
$5,430
$7,407
$9,444
$11,537
$13,703
$15,866
$18,040
$20,265
$22,517
$24,828
$27,131
$29,435
$31,690
$33,909
$36,178
$38,457
$40,766
$43,150
$45,595
$48,143
$50,778
$53,537
$56,398
$59,405
$62,541
$65,846
$69,315
$72,968
$76,809
$80,851
$85,101
$89,571
$94,267
$99,201
$104,385
$761,400
$314,200
           EPA also conducted a separate analysis of the GHG benefits over the model year
    lifetimes of the 2012 through 2016 model year vehicles. In contrast to the calendar year
                                             7-49

-------
Draft Regulatory Impact Analysis

analysis, the model year lifetime analysis shows the lifetime impacts of the program on each
of these MY fleets over the course of its lifetime.  Full details of the inputs to this analysis can
be found in DRIA chapter 5.  The GHG benefits of the full life of each of the five model years
from 2012 through 2016 are shown in Table 7-12 through Table 7-16 for each of the five
different social cost of carbon values. The GHG benefits are shown for each year in the
model year life and in net present value using both a 3 percent and a 7 percent discount rate.
                                        7-50

-------
                                                      Environmental and Health Impacts
Table 7-12: Upstream and Downstream CO2-Equivalent Benefits for the 5% SCC Value, Model Year Analysis
                                   (Mfflions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV,
3%
NPV,
7%
2012
$42
$42
$42
$42
$41
$40
$39
$37
$36
$34
$32
$29
$25
$21
$18
$15
$12
$10
$8
$7
$5
$5
$4
$3
$3
$2
$2
$1
$1
$1
$1
$1
$1
$1
$1
$1
$0
$0
$0
$477
$368
2013
$0
$67
$66
$66
$66
$65
$63
$61
$59
$56
$53
$50
$46
$40
$34
$28
$23
$19
$16
$13
$11
$9
$8
$6
$5
$5
$4
$3
$3
$2
$2
$2
$2
$2
$1
$1
$1
$0
$0
$733
$543
2014
$0
$0
$95
$95
$95
$94
$92
$90
$88
$84
$80
$76
$71
$65
$57
$48
$40
$34
$28
$23
$19
$16
$13
$11
$9
$8
$7
$6
$4
$4
$4
$3
$3
$3
$2
$2
$2
$2
$0
$1,021
$727
2015
$0
$0
$0
$137
$137
$136
$135
$133
$130
$126
$121
$115
$109
$102
$94
$81
$69
$58
$49
$40
$33
$28
$23
$19
$16
$14
$12
$10
$9
$6
$6
$5
$5
$4
$4
$3
$3
$3
$2
$1,426
$978
2016
$0
$0
$0
$0
$188
$187
$187
$185
$182
$178
$172
$166
$158
$150
$140
$128
$112
$95
$80
$67
$56
$46
$38
$31
$26
$22
$19
$16
$14
$12
$9
$8
$7
$6
$6
$5
$5
$4
$4
$1,897
$1,251
SUM
$42
$109
$204
$340
$526
$522
$516
$507
$494
$478
$459
$436
$409
$378
$341
$300
$257
$216
$181
$150
$124
$103
$85
$71
$60
$51
$43
$37
$31
$26
$21
$19
$17
$15
$14
$12
$11
$8
$6
$5,555
$3,866
                                           7-51

-------
    Draft Regulatory Impact Analysis
Table 7-13: Upstream and Downstream CO2-Equivalent Benefits for the 5% Newell-Pizer SCC Value, Model Year
                                 Analysis (Mfflions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV,
3%
NPV,
7%
2012
$84
$84
$84
$83
$82
$80
$78
$75
$71
$68
$64
$58
$50
$42
$35
$29
$24
$20
$16
$13
$11
$9
$8
$7
$5
$5
$3
$3
$3
$2
$2
$2
$2
$2
$1
$1
$0
$0
$0
$955
$736
2013
$0
$133
$133
$133
$131
$129
$126
$122
$118
$112
$107
$100
$91
$79
$67
$56
$47
$39
$32
$26
$22
$18
$15
$13
$11
$9
$8
$6
$5
$5
$4
$4
$3
$3
$3
$2
$2
$0
$0
$1,467
$1,086
2014
$0
$0
$191
$190
$190
$187
$185
$180
$175
$168
$160
$152
$142
$130
$113
$96
$81
$68
$56
$47
$38
$32
$26
$22
$19
$16
$14
$12
$9
$8
$7
$6
$6
$5
$5
$4
$4
$3
$0
$2,042
$1,453
2015
$0
$0
$0
$274
$273
$273
$270
$266
$260
$252
$242
$231
$219
$205
$187
$163
$139
$116
$97
$81
$67
$55
$46
$38
$32
$27
$23
$20
$17
$13
$12
$10
$9
$8
$7
$7
$6
$5
$5
$2,853
$1,955
2016
$0
$0
$0
$0
$376
$374
$374
$370
$364
$356
$345
$332
$316
$300
$280
$256
$223
$190
$160
$133
$111
$92
$76
$63
$52
$44
$37
$32
$28
$24
$18
$16
$14
$13
$12
$10
$9
$8
$7
$3,794
$2,503
SUM
$84
$217
$407
$680
$1,052
$1,044
$1,033
$1,013
$988
$956
$918
$872
$818
$756
$683
$601
$513
$432
$361
$300
$249
$206
$171
$142
$120
$102
$86
$73
$62
$52
$43
$38
$35
$31
$28
$25
$21
$17
$12
$11,109
$7,733
                                              7-52

-------
                                                        Environmental and Health Impacts
Table 7-14: Upstream and Downstream CO2-Equivalent Benefits for the from 3% and 5% SCC Value, Model Year
                                 Analysis (Mfflions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV,
3%
NPV,
7%
2012
$160
$160
$160
$158
$156
$152
$148
$142
$136
$129
$121
$110
$95
$81
$67
$56
$46
$38
$31
$25
$21
$17
$14
$12
$10
$9
$6
$6
$5
$4
$4
$4
$3
$3
$3
$2
$0
$0
$0
$1,814
$1,398
2013
$0
$253
$252
$252
$249
$245
$240
$233
$224
$213
$203
$190
$173
$150
$128
$107
$89
$74
$61
$50
$41
$34
$29
$24
$21
$18
$15
$11
$10
$9
$8
$7
$6
$6
$5
$5
$4
$0
$0
$2,786
$2,063
2014
$0
$0
$362
$361
$360
$356
$351
$343
$333
$320
$304
$289
$270
$247
$215
$183
$154
$128
$107
$88
$73
$60
$50
$42
$36
$31
$26
$23
$17
$15
$14
$12
$11
$10
$9
$8
$7
$6
$0
$3,879
$2,762
2015
$0
$0
$0
$521
$519
$519
$513
$505
$493
$479
$460
$438
$416
$389
$355
$309
$263
$221
$185
$154
$127
$105
$87
$72
$61
$52
$45
$38
$33
$24
$22
$20
$18
$16
$14
$13
$11
$10
$9
$5,420
$3,715
2016
$0
$0
$0
$0
$714
$711
$711
$702
$692
$676
$655
$630
$600
$569
$532
$487
$424
$361
$303
$253
$211
$175
$144
$119
$99
$84
$71
$62
$53
$46
$34
$30
$27
$25
$22
$20
$18
$16
$14
$7,208
$4,756
SUM
$160
$413
$774
$1,292
$1,998
$1,984
$1,962
$1,925
$1,878
$1,816
$1,743
$1,657
$1,555
$1,436
$1,298
$1,141
$976
$821
$686
$570
$473
$391
$324
$270
$227
$193
$164
$139
$118
$99
$81
$73
$66
$59
$53
$47
$40
$32
$23
$21,108
$14,693
                                              7-53

-------
  Draft Regulatory Impact Analysis
Table 7-15: Upstream and Downstream COl-Equivalent Benefits for the 3% SCC Value, Model Year Analysis
                                   (Mfflions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV,
3%
NPV,
7%
2012
$278
$277
$277
$274
$270
$264
$257
$247
$236
$224
$210
$191
$166
$140
$117
$97
$80
$65
$53
$43
$36
$30
$25
$21
$18
$16
$11
$10
$9
$8
$7
$6
$6
$5
$4
$4
$0
$0
$0
$3,151
$2,428
2013
$0
$439
$438
$437
$432
$426
$416
$404
$389
$371
$352
$329
$301
$261
$222
$186
$154
$128
$105
$87
$71
$59
$50
$42
$36
$31
$27
$19
$17
$15
$14
$12
$11
$10
$9
$8
$7
$0
$0
$4,840
$3,584
2014
$0
$0
$629
$627
$626
$619
$610
$595
$578
$555
$529
$502
$469
$429
$373
$318
$267
$223
$186
$153
$127
$105
$87
$74
$62
$54
$46
$40
$29
$26
$24
$21
$19
$17
$15
$14
$12
$11
$0
$6,738
$4,796
2015
$0
$0
$0
$905
$902
$901
$890
$878
$857
$831
$799
$761
$722
$676
$617
$537
$457
$384
$321
$267
$221
$182
$151
$125
$106
$90
$78
$66
$58
$42
$38
$34
$31
$27
$25
$22
$20
$18
$16
$9,414
$6,452
2016
$0
$0
$0
$0
$1,240
$1,235
$1,234
$1,220
$1,202
$1,174
$1,138
$1,094
$1,042
$989
$925
$845
$736
$627
$527
$440
$367
$303
$250
$207
$173
$146
$124
$107
$91
$80
$59
$53
$47
$43
$38
$34
$31
$27
$24
$12,519
$8,260
SUM
$278
$717
$1,344
$2,243
$3,471
$3,445
$3,407
$3,344
$3,261
$3,154
$3,028
$2,878
$2,700
$2,494
$2,254
$1,982
$1,694
$1,427
$1,192
$990
$821
$679
$563
$469
$395
$336
$284
$242
$204
$172
$141
$127
$114
$102
$91
$82
$70
$56
$40
$36,661
$25,519
                                            7-54

-------
                                                        Environmental and Health Impacts
Table 7-16: Upstream and Downstream CO2-Equivalent Benefits for the 3% Newell-Pizer SCC Value, Model Year
                                 Analysis (Mfflions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV,
3%
NPV,
7%
2012
$463
$462
$462
$457
$451
$441
$428
$412
$393
$373
$350
$319
$276
$233
$195
$161
$133
$109
$89
$72
$60
$50
$42
$36
$30
$26
$18
$16
$14
$13
$12
$10
$9
$8
$7
$7
$0
$0
$0
$5,251
$4,046
2013
$0
$732
$730
$729
$721
$711
$694
$674
$648
$618
$586
$549
$501
$435
$369
$309
$257
$213
$176
$144
$119
$98
$83
$70
$60
$51
$44
$32
$29
$26
$23
$21
$18
$17
$15
$13
$12
$0
$0
$8,066
$5,973
2014
$0
$0
$1,048
$1,044
$1,043
$1,031
$1,016
$992
$963
$925
$881
$836
$782
$715
$622
$530
$445
$371
$309
$256
$211
$174
$145
$123
$104
$90
$77
$67
$49
$44
$40
$36
$32
$29
$26
$23
$20
$18
$0
$11,229
$7,994
2015
$0
$0
$0
$1,508
$1,503
$1,501
$1,484
$1,463
$1,428
$1,385
$1,332
$1,269
$1,204
$1,126
$1,029
$896
$762
$640
$534
$445
$368
$303
$251
$209
$177
$149
$129
$110
$96
$70
$63
$57
$51
$46
$41
$37
$33
$29
$26
$15,690
$10,753
2016
$0
$0
$0
$0
$2,067
$2,059
$2,057
$2,033
$2,003
$1,956
$1,897
$1,824
$1,737
$1,648
$1,541
$1,408
$1,227
$1,044
$878
$733
$611
$506
$417
$345
$288
$243
$206
$178
$152
$133
$98
$88
$79
$71
$64
$57
$51
$46
$41
$20,865
$13,766
SUM
$463
$1,194
$2,240
$3,739
$5,784
$5,742
$5,679
$5,573
$5,435
$5,257
$5,046
$4,796
$4,500
$4,157
$3,756
$3,304
$2,824
$2,378
$1,986
$1,651
$1,368
$1,132
$938
$782
$658
$560
$474
$403
$340
$286
$236
$212
$190
$170
$152
$137
$116
$93
$67
$61,102
$42,531
                                              7-55

-------
Draft Regulatory Impact Analysis
7.6 Weight Reduction and Vehicle Safety

        Over the past 20 years there has been a generally increasing trend in the weight of
vehicles (see figure III.X-1 below from EPA's Fuel Economy Trends Report).289  There have
been a number of factors contributing to this including: greater penetration of heavier trucks,
introduction of SUVs, and an increasing amount of content in vehicles (including features for
safety, noise reduction, added comfort, luxury, etc). This increased weight has been partially
enabled by the increased efficiency of vehicles, especially in engines and transmissions.  The
impressive improvements in efficiency during this period have not only allowed for greater
weight carrying capacity (and towing), but it has also allowed for greater acceleration
performance in the fleet. Unfortunately, as the figure also shows, none of this efficiency
improvement has been realized in fuel economy gains or GHG emissions reductions.
            Fuel Economy and Performance
             (Three Year Moving Average)
                       Cars
Fuel Economy and Performance
 (Three Year Moving Average)
          Wagons
            MPG, 0 to 60 (sec.)
          30-	4500
                           Inertia Wflgnt fasj ,
                                                MPG, 0 to 60 (sec.)
             0 to 60 Tlme~~~
                                                               Inertia Wflgnt (its J
           1975  1980 19&5 1990 1995 ZOOO 2005
                                               1915  1MO  1965 1990 1995 2000 2005
  Figure 7-6: Weight. O-to-60 MPH acceleration time and adjusted fuel economy for light-duty vehicles

       During this same period, the safety of vehicles has also undergone tremendous
improvement.  Vehicles are designed to better withstand both frontal and side impacts,
occupants are protected better with increased seat belt usage and air bags, and drivers are able
to avoid accidents with anti-lock brakes (ABS), electronic stability control (ESC), and
improved tires and suspension. NHTSA anticipates a 12.6 percent reduction in fatality levels
between 2007 and 2020 with safety improvements due to pending NHTSA FMVSS and other
factors. Assuming that safety improvements will be made evenly throughout that period, EPA
estimates the reduction in fatalities between 2007 and 2016 to be 8.7%.

       The interplay between vehicle weight and potential impact on safety is complex.
While certainly an effective option for reducing CCh emissions, the reduction of vehicle
weight is a controversial and complicated topic. In a joint technical analysis, EPA and
NHTSA agree that automakers could reduce weight as one part of the industries' strategy for
                                         7-56

-------
                                                  Environmental and Health Impacts

meeting the proposed standards.  As shown in table III.D.6-3 of the Preamble, EPA is
expecting that vehicle manufacturers will reduce the weight of their vehicles by
approximately 4% on average between 2011 and 2016 although individual vehicles may have
greater or smaller weight reduction (NHTSA's results are similar using the Volpe model.)
The penetration and magnitude of these changes are consistent with the public announcements
made by many of the manufacturers since early 2008 and are consistent with the meetings that
EPA has had with senior engineers and technical leadership at many of the automotive
companies during 2008 and 2009.

       Between September 2008 and May  2009, EPA met with 11 major auto companies:
GM, Chrysler, Ford, Nissan, Honda, Toyota, Mitsubishi, Hyundai/Kia, BMW, Mercedes and
Volkswagen. Each company announced plans to reduce vehicle weight broadly across the
passenger car vehicle and light truck categories within the 2012 to  2016 timeframe. Their
plans for vehicle weight reduction are not limited to a single weight class but instead are
expected to be implemented widely across the their products. The following statements
summarize a number of automotive manufacturers' future plans to reduce vehicle weight
announced in the public domain within the past two year:

   •   Ford: 250 to 750 pound weight reductions 2012 to 2020 across all vehicle platforms

   •   Toyota: 30% weight reduction on 2015 Corolla and a 10% weight reduction on mid-
          size vehicles by 2015

   •   Nissan: 15% average weight reduction by 2015

   •   Mazda: 100 kg (220 pound) weight reduction by  2011 and an additional 100 kg
          weight reduction by 2016

   •   Mercedes: 5% average weight reduction by 2015

       Reducing vehicle mass without reducing the size, footprint or the structural integrity
of the vehicle is technically feasible. Many of the technical options for doing so are outlined
in Chapter 3 of the joint TSD and in this DRIA. Weight reduction can be accomplished by
the proven methods described below. Every manufacturer can employ these methodologies to
some degree, the magnitude to which each will be used will depend on opportunities within
individual vehicle design.

     • Material Substitution:  Substitution of lower density and/or higher strength materials
       in a manner that preserves or improves the function of the component. This includes
       substitution of high-strength steels,  aluminum, magnesium or composite materials for
       components currently fabricated from mild steel, e.g., the magnesium-alloy front
       structure used on the 2009 Ford F150 pickups (we note that since these MY 2009
       F150s have  only begun to enter the fleet, there is little real-world crash data available
       to evaluate the safety impacts of this new design). Light-weight materials with
       acceptable energy absorption properties can maintain structural integrity and
       absorption of crash energy relative to previous designs while providing a net decrease
       in component weight.

                                        7-57

-------
Draft Regulatory Impact Analysis

     • Smart Design:  Computer aided engineering (CAE) tools can be used to better
       optimize load paths within structures by reducing stresses and bending moments
       without adversely affecting structural integrity. 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 with little compromise in vehicle functionality.

     • Reduced Powertrain Requirements: Reducing vehicle weight sufficiently allows for
       the use of a smaller, lighter and more efficient engine while maintaining or increasing
       performance. Approximately  half of the reduction is due to these reduced powertrain
       output requirements from reduced engine power output and/or displacement, lighter
       weight transmission and final drive gear ratios. The subsequent reduced rotating mass
       (e.g. transmission, driveshafts/halfshafts, wheels and tires) via weight and/or size
       reduction of components are made possible by reduced torque output requirements.

     • Mass Decompounding: Following from the point above, the compounded weight
       reductions of the body, engine and drivetrain can reduce stresses on the suspension
       components, steering components, brakes,  and thus allow further reductions in the
       weight of these subsystems.  The reductions in weight for unsprung masses such as
       brakes, control arms, wheels and tires can further reduce stresses in the suspension
       mounting points which can allow  still further reductions in weight.  For example,
       lightweighting can allow for the reduction in the size of the vehicle brake system,
       while  maintaining the same stopping distance. It is estimated that 1.25 kilograms of
       secondary weight savings can be achieved for every kilogram of weight saved on a
       vehicle when all subsystems are redesigned to take into account the initial primary
         •  i  .   •     290
       weight savings.

       Weight reduction is broadly applicable  across all vehicle subsystems including the
engine, exhaust system, transmission, chassis, suspension, brakes, body, closure panels,
glazing, seats and other interior components, engine cooling systems and HVAC systems.
EPA believes it is both technically feasible to reduce weight without reducing vehicle size,
footprint or structural strength and manufacturers have indicated to the agencies that they will
use these approaches to accomplish these tasks. We request written comment on this
assessment and this projection, including up-to-date plans regarding the extent of use by each
manufacturer of each of the methodologies described above.

       EPA also projects that automakers will  not reduce footprint in order to meet the
proposed CCh standards in our modeling analysis.  NHTSA and EPA have  taken two
measures to help ensure that the proposed rules provide no incentive for mass  reduction to be
accompanied by a corresponding decrease in the footprint of the vehicle (with its concomitant
decrease in crush and crumple zones).  The first design feature of the proposed rule is that the
CO2 or fuel economy targets are based  on the attribute of footprint (which is a surrogate for
                                        7-58

-------
                                                    Environmental and Health Impacts

vehicle size).R  The second design feature is that the shape of the footprint curve (or function)
has been carefully chosen such that it neither encourages manufacturers to increase, nor
decrease the footprint of their fleet.  Thus, the standard curves are designed to be
approximately "footprint neutral" within the sloped portion of the function.8  For further
discussion on this, refer to Section II.C of the preamble, or Chapter 2 of the joint TSD. Thus
the agencies are assuming in their modeling analysis that the manufacturers could reduce
vehicle mass without reducing vehicle footprint as one way to respond to the proposed rule.1

       In Section IV of the preamble, NHTS A presents a safety analysis of the proposed
CAFE standards based on the 2003 Kahane analysis.  NHTSA's Dr. Charles Kahane
performed a thorough review on historical data regarding the relationship between mass
reduction, wheel base, track width and fatality risk.291'292  The results from 1991-1999 vehicle
data indicate that a heavier vehicle is safer than a lighter one based on the assumption that
historical vehicle mass reductions are accompanied with vehicle size and footprint reductions.

       As discussed in Section IV of the Preamble, NHTSA has developed a worse case
estimate  of the impact of weight reductions on fatalities. The underlying data used for that
analysis does not allow NHTSA to analyze  the specific impact of weight reduction at constant
footprint because historically there have not been a large number of vehicles produced that
relied substantially on material substitution. Rather, the data set includes vehicles that were
either smaller and lighter or larger and heavier.  The numbers in the NHTSA analysis predict
the safety-related fatality consequences that would occur in the unlikely event that weight
reduction for model years 2012-2016 is accomplished by reducing mass and reducing
footprint. EPA concurs with NHTSA that the safety analysis conducted by NHTSA and
presented in Section IV is a worst case analysis for fatalities and we expect the actual impacts
on vehicle safety  could be much less. EPA and NHTSA are not able to quantify the lower-
bound or the best-case potential impacts at this time.

       The 2005  Dynamic Research, Incorporated (DRI)  studies assessed the independent
effects of vehicle weight and size on safety  in order to determine if there are tradeoffs
between improving vehicle safety and fuel consumption.  In their 2005 studies,293'294 one of
which was published as a Society of Automotive Engineers Technical Paper and received peer
review through that body, DRI presented results that indicate that vehicle weight reduction
tends to decrease fatalities, but vehicle wheelbase and track reduction tends to increase
fatalities. The DRI work focused on four major points, with #1 and #4 being discussed with
additional detail below:
R As the footprint attribute is defined as wheelbase times track width, the footprint target curves do not
discourage manufacturers from reducing vehicle size by reducing front, rear, or side overhang, which can impact
safety by resulting in less crush space.

s This neutrality with respect to footprint does not extend to the smallest and largest vehicles, because the
function is limited, or flattened, in these footprint ranges

T See Chapter 1 of the joint TSD for a description of potential footprint changes in the 2016 reference fleet
                                          7-59

-------
Draft Regulatory Impact Analysis

              1.     2-Door vehicles represented a significant portion of the light duty fleet
       and should not be ignored.

              2.     Directional control and therefore crash avoidance improves with a
       reduction in curb weight.

              3.     The occupants of the impacted vehicle, or "collision partner" benefit
       from being impacted by a lighter vehicle.

              4.     Rollover fatalities are reduced by a reduction in curb weight due to
       lower centers of gravity and lower loads on the roof structures.

       The data used for the DRI analysis was similar to NHTSA's 2003 Kahane study, using
Fatality Analysis Reporting System (PARS) data for vehicle model years 1985 through 1998
for cars, and 1985 through 1997 trucks.  This data overlaps Kahane's PARS data on model
year 1991 to 1999 vehicles. DRI also used a logistical regression method similar to the
approach taken by the 2003 Kahane study.  However, DRI included 2-door passenger cars,
whereas the Kahane study  excluded all 2-door vehicles. The 2003 Kahane study excluded 2-
door passenger cars because it found that for MY 1991-1999 vehicles, sports and muscle cars
constituted a significant proportion of those vehicles. These vehicles have relatively high
weight relative to their wheelbase, and are also disproportionately involved in crashes. Thus,
Kahane concluded that including these vehicles in the analysis excessively skewed the
regression results. As of July 1, 1999, 2-door passenger cars represented 29% of the
registered cars in the United States. The majority of 2-door vehicles excluded in the 2003
Kahane study and included in DRI's analysis were high-sales volume light-duty vehicles and
vehicles shared common vehicle platforms and architectures with 4-door vehicles that were
included in the 2003 Kahane study. Specific examples include the Chevrolet Cavalier and
Monte Carlo, Oldsmobile Achieva and Supreme, Buick Riviera, Ford Escort and Probe,
Mercury Tracer, Honda Civic, Hyundai Accent, and VW Golf which do not necessarily
represent high-weight, short-wheelbase sports and high-performance vehicle types. DRI's
position was that this is a significant portion of the light duty fleet, too large to be ignored,
and conclusions regarding the effects of weight and safety should be based on data for all
cars, not just 4-doors.

       DRI did, however,  state in their conclusions that the results are sensitive to removing
data for 2-doors and wagons, and that the results for 4-door cars with respect to the effects of
wheelbase and track width were no longer statistically significant when 2-door cars were
removed.  EPA and NHTSA recognize the technical challenges of properly accounting for 2-
door cars in a regression analysis evaluating the impacts of vehicle weight on safety, due to
the concerns discussed for the Kahane study above. Thus, the agencies seek comment on how
to ensure that any analysis supporting the final rule accounts as fully as possible for the range
of safety impacts due to weight reduction on the variety of vehicles regulated under these
proposed standards.

       The DRI and Kahane studies also differ with respect to the impact of vehicle weight
on rollover fatalities. The Kahane study treated curb weight as a surrogate for size and weight
and analyzed them as a single variable. Using this method, the 2003 Kahane analysis

                                        7-60

-------
                                                   Environmental and Health Impacts

indicates that curb weight reductions would increase fatalities due to rollovers. The DRI
study differed by analyzing curb weight, wheelbase, and track as multiple variables and
concluded that curb weight reduction would decrease rollover fatalities, and wheelbase and
track reduction would increase rollover fatalities.  DRI offers two potential root causes for
higher curb weight resulting in higher rollover fatalities. The first is that a taller vehicle tends
to be heavier than a shorter vehicle; therefore heavier vehicles may be more likely to rollover
because the vehicle height and weight are correlated with vehicle center of gravity height.
The  second is that FMVSS 216 for roof crush strength requirements for passenger cars of
model years 1995 through 1999 were proportional to the unloaded vehicle weight if the
weight is less than 3,333 Ibs, however they were a constant if the weight is greater than 3,333
Ibs. Therefore heavier vehicles may have had relatively less rollover crashworthiness.

       NHTSA has rejected the DRI analysis, and has not relied on it for its evaluation of
safety impact changes in CAFE standards.  See Section IV.G.6  of this Notice, as well as
NHTSAs March 2009 Final Rulemaking for MY2011 CAFE standards (see 74 FR at 14402-
05).

       The DRI and Kahane analysis of the PARS data appear  to be quite similar in one
respect because the results are reproducible between the two studies when using aggregated
vehicle attributes for 4-door cars.293'294'295  The two analyses differ when individual vehicle
attributes of mass, wheelbase and track width are separately analyzed.  NHTSA has raised this
as a concern with the DRI  study.   When 2-door vehicles are removed from the data set EPA
is concerned that the results may no longer be statistically significant with respect to
independent vehicle attributes due to the small size of the remaining data set, as DRI stated in
the 2005 study.

       The DRI analysis concluded that there would be small additional reductions in
fatalities for cars and trucks if the weight reduction occurs  without accompanying vehicle
footprint or size changes.  EPA notes that if DRFs results were  to be applied using the curb
weight reductions predicted by the OMEGA model, an overall reduction in fatalities would be
predicted. EPA invites comment on all aspects of the issue of the impact of this kind of
weight reduction on safety, including the usefulness of the DRI study in evaluating this issue.

       The agencies are committed to continuing  to analyze vehicle safety issues so a more
informed evaluation can be made.  We request comment on this issue.  These comments
should include not only further discussion and analysis of the relevant studies but  data and
analysis which can allow the agencies to more accurately quantify any potential safety issues
with the proposed standards.
                                         7-61

-------
Draft Regulatory Impact Analysis
        References

        References can be found in the EPA DOCKET: EPA-HQ-OAR-2009-0472 or are
        publically available.
 U.S. EPA. (2005). Review of the National Ambient Air Quality Standard for P articulate Matter: Policy
Assessment of Scientific and Technical Information, OAQPS Staff Paper.  EPA-452/R-05-005a. Retrieved
March 19, 2009 from http://www.epa.gov/ttn/naaqs/standards/pm/data/pmstaffpaper_20051221.pdf. Section 2.2.

2 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF.  Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.

3U.S. EPA. (2005). Review of the National Ambient Air Quality Standard for Particulate Matter: Policy
Assessment of Scientific and Technical Information, OAQPS Staff Paper.  EPA-452/R-05-005a. Retrieved
March 19, 2009 from http://www.epa.gov/ttn/naaqs/standards/pm/data/pmstaffpaper_20051221.pdf.

4 U.S. EPA. (2006). Provisional Assessment of'Recent Studies on Health Effects of Particulate Matter Exposure.
EPA/600/R-06/063. Retrieved on March 19, 2009 from
http://www.epa.gov/air/particlepollution/pdfs/ord_report_20060720.pdf.

5 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.  p. 8-305.

6 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.  p. 9-93.

7 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   Section 8.3.3.1.

8 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   Table 8-34.

9 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   Section 8.3.1.3.4.

10 U.S. EPA.  (2006). National Ambient Air Quality Standards for Particulate Matter, Proposed Rule. 71 FR
2620, January 17, 2006.

11 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   Section 8.3.4.

12 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   p. 8-85.

13 Laden, F., Neas, L.M., Dockery D.W., et al. (2000). Association of fine particulate matter from different
sources with  daily mortality in sixU.S. cities. Environ Health Perspectives, 108(10), 941-947.


                                              7-62

-------
                                                            Environmental and Health Impacts
14 Schwartz, J., Laden, F. Zanobetti, A. (2002). The concentration-response relation between PM(2.5) and daily
deaths. Environ Health Perspect, 110(10), 1025-1029.

15 Mar, T.F., Ito, K., Koenig, J.Q, Larson, T.V., Eatough, D.J., Henry, R.C., Kim, E., Laden, F., Lall, R, Neas,
L., Stolzel, M., Paatero, P., Hopke, P.K., Thurston, G.D. (2006). PM source apportionment and health effects. 3.
Investigation of inter-method variations in associations between estimated source contributions of PM2.5 and
daily mortality in Phoenix, AZ. J. Exposure Anal. Environ. Epidemiol, 16, 311-320.

16 Ito, K., Christensen, W.F., Eatough, D.J., Henry, R.C., Kim, E., Laden, F., Lall, R., Larson, T.V., Neas, L.,
Hopke, P.K., Thurston, G.D. (2006). PM source apportionment and health effects: 2. An investigation of
intermethod variability in associations between source-apportioned fine particle mass and daily mortality in
Washington, DC. J. Exposure Anal. Environ. Epidemiol., 16, 300-310.

17 Janssen N.A., Schwartz J., Zanobetti A., et al. (2002). Air conditioning and source-specific particles as
modifiers of the effect of PM10 on hospital admissions for heart and lung disease. Environ Health Perspect,
110(1), 43-49.

18 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume  I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   p. 8-307.

19 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume  I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   p. 8-313, 8-314.

20 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume  I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   p.8-318.

21 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume  I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   p. 8-306.

22 U.S. EPA. (2005). Review of the National Ambient Air Quality Standard for P articulate Matter: Policy
Assessment of Scientific andTechnical Information,  OAQPS Staff Paper.  EPA-452/R-05-005a. Retrieved
March 19, 2009 from http://www.epa.gov/ttn/naaqs/standards/pm/data/pmstaffpaper_20051221.pdf p.3-18.

23 Dockery, D.W., Pope, CA. Ill, Xu, X, et al. (1993). An association between air pollution and mortality in six
U.S. cities. NEnglJMed, 529,1753-1759. Retrieved on March 19, 2009 from
http://content.nejm.org/cgi/content/full/329/24/1753.

24 Pope, C.A., III, Thun, M.J., Namboodiri, M.M., Dockery, D.W., Evans, J.S., Speizer, F.E., and Heath, C.W.,
Jr. (1995). Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. Am. J. Respir.
Crit. Care Med, 151, 669-674.

25 Pope, C. A., Ill, Burnett, R.T., Thun, M. J., Calle,  E.E., Krewski, D., Ito, K., Thurston, G.D., (2002). Lung
cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. J. Am. Med. Assoc,
257,1132-1141.

26 Krewski, D., Burnett, R.T., Goldberg, M.S., et al. (2000). Reanalysis of the Harvard Six Cities study and the
American Cancer Society study of particulate air pollution andmortality. A special report of the Institute's
Particle Epidemiology Reanalysis Project.  Cambridge, MA: Health Effects Institute. Retrieved on March 19,
2009 from http://es.epa.gov/ncer/science/pm/hei/Rean-ExecSumm.pdf

27 Jerrett, M., Burnett, R.T., Ma, R., et al. (2005). Spatial Analysis of Air Pollution and Mortality in Los Angeles.
Epidemiology, 16(6),121-136.

28 U.S. EPA (2004). Air Quality Criteria for Particulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.   Section 9.2.2.1.2.
                                                7-63

-------
Draft Regulatory Impact Analysis
29 Kiinzli, N., Jerrett, M., Mack, W.J., et al. (2004). Ambient air pollution and atherosclerosis in Los Angeles.
Environ Health Perspect. ,113,201-206

30 U.S. EPA. (2006J. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

31 U.S. EPA. (2007). Review of the National Ambient Air Quality Standards for Ozone: Policy Assessment of
Scientific and Technical Information, OAQPS Staff Paper. EPA-452/R-07-003. Washington, DC, U.S. EPA.
Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at http://www.regulations.gov/.

32National Research Council (NRC), 2008. Estimating Mortality Risk Reduction and Economic Benefits from
Controlling Ozone Air Pollution. The National Academies Press: Washington, D.C.

33 Bates, D.V., Baker-Anderson, M.,  Sizto, R. (1990).  Asthma attack periodicity: a study of hospital emergency
visits  in Vancouver. Environ. Res., 51,51-70.

34 Thurston, G.D., Ito, K., Kinney, P.L., Lippmann, M. (1992). A multi-year study of air pollution and
respiratory hospital admissions in three New York State metropolitan areas:  results for 1988 and 1989 summers.
J. Exposure Anal. Environ. Epidemiol, 2,429-450.

35 Thurston, G.D., Ito, K., Hayes, C.G., Bates, D.V., Lippmann, M. (1994) Respiratory hospital admissions and
summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols. Environ. Res., 65,
271-290.

36 Lipfert, F.W., Hammerstrom, T. (1992). Temporal patterns in air pollution and hospital admissions. Environ.
Res., 59,374-399.

37 Burnett, R.T., Dales, R.E., Raizenne, M.E., Krewski, D., Summers, P.W., Roberts, G.R., Raad-Young, M.,
Dann,T., Brook, J. (1994). Effects of low ambient levels of ozone and sulfates on the frequency of respiratory
admissions to Ontario hospitals. Environ. Res., 65, 172-194.

38 U.S. EPA. (2006J. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

39 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

40 Devlin, R. B., McDonnell, W. F., Mann, R., Becker, S., House, D. E., Schreinemachers, D., Koren, H. S.
(1991). Exposure of humans to ambient levels of ozone for 6.6 hours causes cellullar and biochemical changes in
the lung. Am. J. Respir.  CellMol. Biol, 4, 72-81.

41 Koren, H. S., Devlin, R. B., Becker, S., Perez, R., McDonnell, W. F. (1991). Time-dependent changes of
markers associated with inflammation in the lungs of humans exposed to ambient levels of ozone. Toxicol.
Pathol, 19, 406-411.

42Koren, H. S., Devlin, R. B., Graham, D. E., Mann, R., McGee, M. P., Horstman, D. H., Kozumbo, W. J.,
Becker, S., House, D. E., McDonnell, W. F., Bromberg, P. A. (1989). Ozone-induced inflammation in the lower
airways of human subjects. Am. Rev. Respir. Dis., 39,  407-415.

43 Schelegle, E.S., Siefkin, A.D., McDonald, R.J. (1991).  Time course of ozone-induced neutrophilia in normal
humans. Am. Rev. Respir. Dis., ^5,1353-1358.

44 U.S. EPA. (1996). Air Quality Criteria for Ozone and Related Photochemical Oxidants. EPA600-P-93-004aF.
Washington. D.C.: U.S. EPA. Retrieved on March 19, 2009 from EPA-HQ-OAR-2005-0161. p. 7-171.

45 Hodgkin,  J.E., Abbey, D.E., Euler, G.L., Magie, A.R. (1984). COPD  prevalence in nonsmokers in high and
low photochemical air pollution areas. Chest, 86, 830-838.
                                               7-64

-------
                                                            Environmental and Health Impacts
46 Euler, G.L., Abbey, D.E., Hodgkin, J.E., Magie, A.R. (1988).  Chronic obstructive pulmonary disease
symptom effects of long-term cumulative exposure to ambient levels of total oxidants and nitrogen dioxide in
California Seventh-day Adventist residents.  Arch. Environ. Health, 43, 279-285.

47 Abbey, D.E., Petersen, F., Mills, P.K., Beeson, W.L. (1993).  Long-term ambient concentrations of total
suspended particulates, ozone, and sulfur dioxide and respiratory symptoms in a nonsmoking population. Arch.
Environ. Health, 48, 33-46.

48 U.S. EPA. (2007). Review of the National Ambient Air Quality Standards for Ozone: Policy Assessment of
Scientific and Technical Information, OAQPS Staff Paper. EPA-452/R-07-003. Washington, DC, U.S. EPA.
Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at http://www.regulations.gov/.

49 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

50 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

51 Avol, E.L., Trim, S.  C., Little, D.E., Spier, C.E., Smith, M. N., Peng, R.-C., Linn, W.S., Hackney, J.D., Gross,
K.B.,  D'Arcy, J.B., Gibbons, D., Higgins, I.T.T. (1990 June). Ozone exposure and lung function in children
attending a southern California summer camp. Paper no. 90-150.3.  Paper presented at the 83rd annual meeting
and exhibition of the Air & Waste Management Association, Pittsburgh, PA.

52 Higgins, I. T.T., D'Arcy, J. B., Gibbons, D. I., Avol, E. L., Gross, K.B. (1990). Effect of exposures to ambient
ozone on ventilatory lung function in children. Am. Rev. Respir. Dis., 141, 1136-1146.

53 Raizenne, M.E., Burnett, R.T., Stern,  B., Franklin,  C.A., Spengler, J.D. (1989) Acute lung function responses
to ambient acid aerosol exposures in children. Environ. Health Perspect., 79,179-185.

54 Raizenne, M.; Stern, B.; Burnett, R.; Spengler, J. (1987 June) Acute respiratory function and transported air
pollutants: observational studies. Paper no. 87-32.6.  Paper presented at the 80th annual meeting of the Air
Pollution Control Association, New York, NY.

55 Spektor, D. M., Lippmann, M. (1991). Health effects of ambient ozone on healthy children at a summer camp.
In: Berglund, R. L.;  Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
international conference', March 1990; Los Angeles,  CA. Pittsburgh, PA: Air & Waste Management
Association; pp. 83-89. (A&WMA transaction series no. TR-19).

56 Spektor, D. M., Thurston, G.D., Mao, J., He, D., Hayes, C., Lippmann, M. (1991). Effects of single- and
multiday ozone exposures on respiratory function in active normal children. Environ. Res, 55,107-122.

57 Spektor, D. M., Lippman, M., Lioy, P. J., Thurston, G. D., Citak,  K., James, D. J., Bock, N., Speizer, F.  E.,
Hayes, C. (1988). Effects of ambient ozone on respiratory function in active, normal children. Am. Rev. Respir.
Dis., ^57,313-320.

58 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

59 Hazucha, M. J., Folinsbee, L. J., Seal, E., Jr. (1992). Effects of steady-state and variable ozone concentration
profiles on pulmonary function. Am. Rev. Respir. Dis., 146, 1487-1493.

60 Horstman, D.H., Ball, B.A., Folinsbee, L.J., Brown, J., Gerrity, T. (1995) Comparison of pulmonary responses
of asthmatic and nonasthmatic subjects  performing light exercise while exposed to a low level of ozone.
Toxicol. Ind. Health, 11(4),  369-85.

61 Horstman, D.H.,; Folinsbee, L.J., Ives, P.J., Abdul-Salaam, S., McDonnell, W.F. (1990). Ozone concentration
and pulmonary response  relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10,
and 0.12 ppm. Am. Rev. Respir. Dis.,  142, 1158-1163.
                                                7-65

-------
Draft Regulatory Impact Analysis
62 U.S. EPA. (2008). Integrated Science Assessment (ISA) for Sulfur Oxides-Health Criteria (Final Report).
EPA/600/R-08/047F. Washington, DC: U.S. Environmental Protection Agency. Retrieved on March 18, 2009
from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l98843

63 U.S. EPA (2008). Integrated Science Assessment for Oxides of Nitrogen -Health Criteria (Final Report).
EPA/600/R-08/071. Washington, DC,: U.S.EPA. Retrieved on March 19, 2009 from
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=194645.

64 U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide. EPA600-P-99-001F. June 1, 2000. U.S.
Environmental Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, D.C. This document is available online at http://www.epa.gov/ncea/pdfs/coaqcd.pdf
A copy of this document is available in Docket EPA-HQ-OAR-2005-0161.

65 U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide. EPA600-P-99-001F. June 1, 2000. U.S.
Environmental Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, D.C. This document is available online at http://www.epa.gov/ncea/pdfs/coaqcd.pdf
A copy of this document is available in Docket EPA-HQ-OAR-2005-0161.

66 Coburn, R.F. (1979) Mechanisms of carbon monoxide toxicity.  Prev.  Med. 8:310-322.

67 Helfaer, M.A., and Traystman, R.J. (1996) Cerebrovascular effects of carbon monoxide.  In: Carbon
Monoxide (Penney, D.G., ed). Boca Raton, CRC Press, 69-86.

68 Benignus, V.A.  (1994) Behavioral effects of carbon monoxide: meta analyses and extrapolations. J. Appl.
Physiol. 76:1310-1316.

69 U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide. EPA600-P-99-001F. June 1, 2000. U.S.
Environmental Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, D.C. This document is available online at http://www.epa.gov/ncea/pdfs/coaqcd.pdf
A copy of this document is available in Docket EPA-HQ-OAR-2005-0161.

70 U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide. EPA600-P-99-001F. June 1, 2000. U.S.
Environmental Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, D.C. This document is available online at http://www.epa.gov/ncea/pdfs/coaqcd.pdf
A copy of this document is available in Docket EPA-HQ-OAR-2005-0161.

71 U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide. EPA600-P-99-001F. June 1, 2000. U.S.
Environmental Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, D.C. This document is available online at http://www.epa.gov/ncea/pdfs/coaqcd.pdf
A copy of this document is available in Docket EPA-HQ-OAR-2005-0161.

72 U.S. EPA (2000). Air Quality Criteria for Carbon Monoxide. EPA600-P-99-001F. June 1, 2000. U.S.
Environmental Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, D.C. This document is available online at http://www.epa.gov/ncea/pdfs/coaqcd.pdf
A copy of this document is available in Docket EPA-HQ-OAR-2005-0161.

73 U. S. EPA. (2009) 2002 National-Scale Air Toxics Assessment.
http: //www. epa.gov/ttn/atw/nata2002/ri sksum. html

74 U.S. EPA (2007) Control of Hazardous Air Pollutants from Mobile Sources. 72 FR 8428; February 26, 2007.

75 U.S. EPA (2003) Integrated Risk Information System File  of Acrolein. National Center for Environmental
Assessment, Office of Research and Development, Washington, D.C. 2003. This material is available
electronically at http://www.epa.gov/iris/subst/0364.htm.

76 U.S. EPA (2009) National-Scale Air Toxics Assessment for 2002.  This material is available electronically at
http: //www. epa.gov/ttn/atw/nata2002/ri sksum. html.

77 U.S. EPA (2009) National-Scale Air Toxics Assessment for 2002. http://www.epa.gov/ttn/atw/nata2002.
                                               7-66

-------
                                                            Environmental and Health Impacts
78 U.S. EPA. 2000. Integrated Risk Information System File for Benzene. This material is available
electronically at: http://www.epa.gov/iris/subst/0276.htm.

79 International Agency for Research on Cancer, IARC monographs on the evaluation of carcinogenic risk of
chemicals to humans, Volume 29, Some industrial chemicals and dyestuffs, International Agency for Research
on Cancer, World Health Organization, Lyon, France, p. 345-389, 1982.

80 Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry, V.A. (1992) Synergistic action of the benzene
metabolite hydroquinone on myelopoietic stimulating activity of granulocyte/macrophage colony-stimulating
factor in vitro, Proc. Natl. Acad. Sci. 89:3691-3695.

81 International Agency for Research on Cancer (IARC).  1987. Monographs on the evaluation of carcinogenic
risk of chemicals to humans, Volume 29, Supplement 7, Some industrial chemicals and dyestuffs, World Health
Organization, Lyon, France.

82 U.S. Department of Health and Human Services National Toxicology Program 11th Report on Carcinogens
available at: http://ntp.niehs.nih.gov/go/16183.

83Aksoy, M.  (1989).  Hematotoxicity and carcinogenicity of benzene. Environ. Health Perspect. 82:193-197.

84 Goldstein, B.D.  (1988). Benzene toxicity.  Occupational medicine.  State of the Art Reviews. 3:541-554.

85 Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E.  Marti, Y.Z. Wang, M. Linet, L.Q. Xi, W. Lu, M.T.
Smith, N. Titenko-Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B. Hayes (1996) Hematotoxicity among
Chinese workers heavily exposed to benzene. Am. J. Ind. Med. 29: 236-246.

86 U.S. EPA 2002 Toxicological Review of Benzene (Noncancer Effects).  Environmental Protection Agency,
Integrated Risk Information System (IRIS), Research and Development, National Center for Environmental
Assessment, Washington DC. This  material is available electronically at http://www.epa.gov/iris/subst/0276.htm.

87Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, B.; Melikian, A.; Eastmond, D.; Rappaport, S.; Li, H.;
Rupa, D.; Suramaya, R.;  Songnian, W.; Huifant,  Y.; Meng, M.; Winnik, M.; Kwok, E.; Li, Y.; Mu, R.; Xu,
B.; Zhang, X.; Li, K. (2003). HEI Report 115, Validation & Evaluation of Biomarkers in Workers Exposed to
Benzene  in China.

88 Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et al. (2002).  Hematological changes among Chinese
workers with a broad range of benzene exposures. Am. J. Industr. Med. 42: 275-285.

89 Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004).  Hematotoxically in Workers Exposed to Low
Levels of Benzene. Science 306: 1774-1776.

90 Turtletaub, K.W. and Mani, C. (2003). Benzene metabolism in rodents at doses relevant to  human exposure
from Urban Air. Research Reports  Health Effect Inst. Report No.113.

91 U.S. EPA. 2002.  Health Assessment of 1,3-Butadiene. Office of Research and Development, National Center
for Environmental Assessment, Washington Office, Washington, DC.  Report No. EPA600-P-98-001F. This
document is available electronically at http://www.epa.gov/iris/supdocs/buta-sup.pdf

92U.S. EPA. 2002  "Full IRIS Summary for 1,3-butadiene (CASRN 106-99-0)" Environmental Protection
Agency, Integrated Risk Information System (IRIS), Research and Development, National Center for
Environmental Assessment, Washington, DC http://www.epa.gov/iris/subst/0139.htm.

93 International Agency for Research on Cancer (IARC) (1999) Monographs on the evaluation of carcinogenic
risk of chemicals to humans, Volume 71, Re-evaluation of some organic chemicals, hydrazine and hydrogen
peroxide  and Volume 97 (in preparation), World Health Organization, Lyon, France.

94 U.S. Department of Health and Human Services National Toxicology Program 11th Report on Carcinogens
available at: http://ntp.niehs.nih.gov/go/16183.

95 Bevan, C.; Stadler, J.C.; Elliot, G.S.; et al. (1996) Subchronic toxicity of 4-vinylcyclohexene in rats and mice
by inhalation. Fundam. Appl. Toxicol. 32:1-10.

96 U.S. EPA. 1987.  Assessment of Health Risks to Garment Workers and Certain Home Residents from


                                               7-67

-------
Draft Regulatory Impact Analysis
Exposure to Formaldehyde, Office of Pesticides and Toxic Substances, April 1987.

"Hauptmann, M..; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2003. Mortality from
lymphohematopoetic malignancies among workers in formaldehyde industries.  Journal of the National Cancer
Institute 95: 1615-1623.

98 Hauptmann, M..; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2004. Mortality from solid cancers
among workers in formaldehyde industries. American Journal of Epidemiology 159: 1117-1130.

99 Beane Freeman, L. E.; Blair, A.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Hoover, R. N.; Hauptmann, M.
2009. Mortality from lymphohematopoietic malignancies among workers in formaldehyde industries: The
National Cancer Institute cohort. J. National Cancer Inst. 101: 751-761.

100 Pinkerton,  L. E. 2004.  Mortality among a cohort of garment workers exposed to formaldehyde: an update.
Occup. Environ. Med. 61: 193-200.

101 Coggon, D, EC Harris, J Poole, KT Palmer. 2003. Extended follow-up of a cohort of British chemical
workers exposed to formaldehyde. J National Cancer Inst. 95:1608-1615.

102Conolly, RB, JS Kimbell, D Janszen, PM Schlosser, D Kalisak, J Preston, and FJ Miller. 2003.  Biologically
motivated computational modeling of formaldehyde carcinogenicity in the F344 rat. Tox Sci 75: 432-447.

103Conolly, RB, JS Kimbell, D Janszen, PM Schlosser, D Kalisak, J Preston, and FJ Miller. 2004. Human
respiratory tract cancer risks of inhaled formaldehyde: Dose-response predictions derived from biologically-
motivated computational modeling of a combined rodent and human dataset. Tox Sci 82: 279-296.

104 Chemical Industry Institute of Toxicology (CUT).1999. Formaldehyde: Hazard characterization and dose-
response assessment for carcinogenicity by the route of inhalation.  CUT, September 28, 1999. Research
Triangle Park, NC.

105 U.S. EPA.  Analysis of the Sensitivity and Uncertainty in 2-Stage Clonal Growth Models for Formaldehyde
with Relevance to Other Biologically-Based Dose Response (BBDR) Models. U.S. Environmental Protection
Agency, Washington, D.C., EPA/600/R-08/103, 2008

106 Subramaniam, R; Chen, C; Crump, K; .et .al. (2008) Uncertainties in biologically-based modeling of
formaldehyde-induced cancer risk: identification of key issues. Risk Anal 28(4):907-923.

107 Subramaniam, R; Chen, C; Crump, K; .et .al. (2007). Uncertainties in the CUT 2-stage model for
formaldehyde-induced nasal cancer in the F344 rat: a limited sensitivity analysis-I. Risk Anal 27:1237

108 Crump, K; Chen, C; Fox, J; .et  .al. (2008) Sensitivity analysis of biologically motivated model for
formaldehyde-induced respiratory cancer in humans. Ann Occup Hyg 52:481-495.

109 International Agency for Research on Cancer (2006) Formaldehyde, 2-Butoxyethanol and 1-tert-
Butoxypropan-2-ol.  Monographs  Volume 88. World Health Organization, Lyon, France.

110 Agency for Toxic Substances and Disease Registry (ATSDR). 1999. Toxicological profile for Formaldehyde.
Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service.
http://www.atsdr.cdc.gov/toxprofiles/tpl 11.html

111 WHO (2002) Concise International Chemical Assessment Document 40:  Formaldehyde.  Published under the
joint sponsorship of the United Nations Environment Programme, the International Labour Organization, and the
World Health Organization, and produced within the framework of the Inter-Organization Programme for the
Sound Management of Chemicals. Geneva.

112U.S. EPA (1988). Integrated Risk Information System File of Acetaldehyde. Research and Development,
National Center for Environmental Assessment, Washington, DC. This material is available electronically at
http://www.epa.gov/iris/subst/0290.htm.


                                               7-68

-------
                                                            Environmental and Health Impacts
113 U.S. Department of Health and Human Services National Toxicology Program 11th Report on Carcinogens
available at: http://ntp.niehs.nih.gov/go/16183.

114 International Agency for Research on Cancer (IARC). 1999. Re-evaluation of some organic chemicals,
hydrazine, and hydrogen peroxide.  IARC Monographs on the Evaluation of Carcinogenic Risk of Chemical to
Humans, Vol 71. Lyon, France.

115 U.S. EPA (1988). Integrated Risk Information System File of Acetaldehyde. This material is available
electronically at http://www.epa.gov/iris/subst/0290.htm.

116 U.S. EPA. 2003. Integrated Risk Information System File of Acrolein. Research and Development, National
Center for Environmental Assessment, Washington, DC. This material is available electronically at
http://www.epa.gov/iris/subst/0364.htm.

117 Appleman, L.M., R.A. Woutersen, and V.J. Feron. (1982). Inhalation toxicity of acetaldehyde in rats. I. Acute
and subacute studies. Toxicology. 23: 293-297.

118 Myou, S.; Fujimura, M.; Nishi K.; Ohka, T.; and Matsuda, T. (1993) Aerosolized acetaldehyde induces
histamine-mediated bronchoconstriction in asthmatics. Am. Rev. Respir.Dis. 148(4 Pt 1): 940-943.

119 Integrated Risk Information System File of Acrolein.  Research and Development, National Center for
Environmental Assessment, Washington, DC. This material is available at http://www.epa.gov/iris/subst/0364.htm

120 International Agency for Research on Cancer (IARC).  1995. Monographs on the evaluation
of carcinogenic risk of chemicals  to humans, Volume 63, Dry cleaning, some chlorinated
solvents and other industrial chemicals, World Health Organization, Lyon, France.

121 Weber-Tschopp, A; Fischer, T; Gierer, R; et al. (1977) Experimentelle reizwirkungen von Acrolein auf den
Menschen. Int Arch Occup Environ Hlth 40(2): 117-130. In German

122 Sim, VM; Pattle, RE. (1957) Effect of possible smog irritants on human subjects. J Am Med Assoc
165(15):1908-1913.

123 Morris JB, Symanowicz PT, Olsen JE, et al. 2003. Immediate sensory nerve-mediated respiratory responses
to irritants in healthy and allergic airway-diseased mice. J Appl Physiol 94(4): 1563-1571.

124 Sim VM, Pattle RE. Effect of possible smog irritants on human subjects JAMA165: 1980-2010, 1957.

125 Perera, F.P.; Rauh, V.; Tsai, W-Y.; et al. (2002) Effect of transplacental exposure to environmental pollutants
on birth outcomes in a multiethnic population. Environ Health Perspect. Ill: 201-205.

126 Perera, F.P.; Rauh, V.; Whyatt, R.M.; Tsai, W.Y.; Tang, D.; Diaz, D.; Hoepner, L.; Barr, D.; Tu, Y.H.;
Camann, D.; Kinney, P. (2006) Effect of prenatal exposure to airborne polycyclic aromatic hydrocarbons on
neurodevelopment in the first 3 years of life among inner-city children. Environ Health Perspect 114: 1287-
1292.

127U. S. EPA.  2004. Toxicological Review of Naphthalene (Reassessment of the Inhalation Cancer Risk),
Environmental Protection Agency, Integrated Risk Information System, Research and Development, National
Center for Environmental Assessment, Washington, DC. This material is available electronically at
http://www.epa.gov/iris/subst/0436.htm.

128 Oak Ridge Institute for  Science and Education. (2004).  External Peer Review for the IRIS Reassessment of
the Inhalation Carcinogenicity of Naphthalene. August 2004.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=84403

129 National Toxicology Program (NTP). (2004). 11th Report on Carcinogens. Public Health Service, U.S.
Department of Health and  Human Services, Research Triangle Park, NC.  Available from: http://ntp-
server.niehs.nih.gov.

130 International Agency for Research on Cancer (IARC). (2002).  Monographs on the Evaluation of the


                                               7-69

-------
Draft Regulatory Impact Analysis
Carcinogenic Risk of Chemicals for Humans. Vol.82. Lyon, France.

131U. S. EPA. 1998. lexicological Review of Naphthalene, Environmental Protection Agency, Integrated Risk
Information System,  Research and Development, National Center for Environmental Assessment, Washington,
DC. This material is available electronically at http://www.epa.gov/iris/subst/0436.htm

132 U.S. EPA Integrated Risk Information System (IRIS) database is available at:  www.epa.gov/iris

133 U.S. EPA (2004).  Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/. p. 4-179.

134 U.S. EPA (2004).  Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/. p. 4-236.

135 U.S. EPA (2004).  Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/. p. 4-182.

136 Sisler, J.F. (1996) Spatial and seasonal patterns and long term variability of the composition of the haze in
the United States: an analysis of data from the IMPROVE network. CIRA Report, ISSN 0737-5352-32, Colorado
State University.

137U.S.EPA. 1999. The Benefits and Costs of the Clean Air Act, 1990-2010.  Prepared for U.S. Congress by
U.S. EPA, Office of Air and Radiation, Office of Policy Analysis and Review, Washington, DC, November;
EPA report no. EPA410-R-99-001.  This document is contained in Docket Identification EPA-HQ-OAR-2004-
0008-0485.

138U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.

139 Winner, W.E., and C.J. Atkinson. 1986.  "Absorption of air pollution by plants, and consequences for growth."
Trends in Ecology and Evolution _/:15-18.

140U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document is contained in Docket Identification EPA-
HQ-OAR-2004-0008-0455 to 0457

141 Tingey, D.T., and  Taylor, G.E. (1982) Variation in plant response to ozone: a conceptual model of
physiological events. In M.H. Unsworth & D.P. Omrod (Eds.), Effects of Gaseous Air Pollution in Agriculture
and Horticulture, (pp.113-138). London, UK: Butterworth Scientific

142U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document is contained in Docket Identification EPA-
HQ-OAR-2004-0008-0455 to 0457

143U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document is contained in Docket Identification EPA-HQ-
OAR-2004-0008-0455 to 0457

144 U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document is contained in Docket Identification EPA-
HQ-OAR-2004-0008-0455 to 0457.

145U.S. E?K. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006. This document is contained in Docket Identification EPA-HQ-
OAR-2004-0008-0455 to 0457

146Ollinger, S.V., Aber, J.D., Reich, P.B. (1997). Simulating ozone effects on forest productivity: interactions
between leaf canopy  and stand level processes. Ecological Applications, 7, 1237-1251.
                                               7-70

-------
                                                            Environmental and Health Impacts
147 Winner, W.E. (1994). Mechanistic analysis of plant responses to air pollution. Ecological Applications, 4(4),
651-661.

148U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.  This document is contained in Docket Identification EPA-
HQ-OAR-2004-0008-0455 to 0457.

149U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.  This document is contained in Docket Identification EPA-
HQ-OAR-2004-0008-0455 to 0457.

150 Fox, S., Mickler, R. A. (Eds.). (1996). Impact of Air Pollutants on Southern Pine Forests, Ecological Studies.
(Vol. 118, 513 pp.)New York: Springer-Verlag.

151 De Steiguer, J.,  Pye, J., Love, C. (1990). Air Pollution Damage to U.S. Forests. Journal of Forestry, 88(8),
17-22.

152 Pye, J.M. (1988). Impact of ozone on the growth and yield of trees: A review. Journal of Environmental
Quality, 17, 347-360.

153U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.

154U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.

155McBride, J.R., Miller, P.R., Laven, R.D. (1985). Effects of oxidant air pollutants on forest succession in the
mixed conifer forest type of southern California. In: Air Pollutants Effects On Forest Ecosystems, Symposium
Proceedings, St. P, 1985, p.  157-167.

156 Miller, P.R., O.C. Taylor, R.G. Wilhour. 1982. Oxidant air pollution effects on a western coniferous forest
ecosystem. Corvallis, OR: U.S. Environmental Protection Agency, Environmental Research Laboratory
(EPA600-D-82-276).

157U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.

158 Kopp, R. J., Vaughn, W. J., Hazilla, M., Carson, R. (1985). Implications of environmental policy for U.S.
agriculture: the case of ambient ozone standards. Journal of Environmental Management, 20, 321-331.

159 Adams, R. M., Hamilton,  S. A., McCarl, B. A. (1986).  The benefits of pollution control: the case of ozone
and U.S. agriculture. American Journal of Agricultural Economics, 34, 3-19.

160 Adams, R. M., Glyer, J. D., Johnson, S. L., McCarl, B. A. (1989). A reassessment of the economic effects of
ozone on U.S. agriculture. Journal of the Air Pollution Control Association, 39,  960-968

161 Abt Associates, Inc. 1995. Urban ornamental plants: sensitivity to ozone and potential economic losses.
U.S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park.  Under contract to RADIAN
Corporation, contract no. 68-D3-0033, WA no. 6. pp. 9-10.

162U.S. EPA. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF, 2006.This document is available in  Docket EPA-HQ-OAR-2005-
0036

163 Grulke, N.E. (2003). The physiological basis of ozone injury assessment attributes  in Sierran conifers.  In A.
Bytnerowicz, M.J. Arbaugh,  & R. Alonso (Eds.), Ozone air pollution in the Sierra Nevada: Distribution and
effects on forests, (pp. 55-81). New York, NY: Elsevier Science, Ltd.

164 White, D., Kimerling, A.J., Overton, W.S. (1992). Cartographic and geometric component of a global
sampling design for environmental monitoring. Cartography and Geographic Information Systems, 19, 5-22.
                                               7-71

-------
Draft Regulatory Impact Analysis
165 Smith, G., Collision, J., Jepsen, E., Prichard, T. (2003). A national ozone biomonitoring program—results
from field surveys of ozone sensitive plants in Northeastern forests (1994-2000). Environmental Monitoring and
Assessment, 87, 271-291.

166 White, D., Kimerling, A.J., Overton, W.S. (1992). Cartographic and geometric component of a global
sampling design for environmental monitoring. Cartography and Geographic Information Systems, 19, 5-22.

167 Smith, G., Coulston, J., Jepsen, E., Prichard, T. (2003). A national ozone biomonitoring program—results
from field surveys of ozone sensitive plants in Northeastern forests (1994-2000). Environmental Monitoring and
Assessment, §7,271-291.

168 Coulston, J.W., Riitters, K.H., Smith, G.C. (2004). A preliminary assessment of the Montreal process indica-
tors of air pollution for the United States. Environmental Monitoring and Assessment, 95, 57-74.

169 U.S. EPA (United States Environmental Protection Agency). 2006. Air quality criteria for ozone and related
photochemical oxidants. EPA/600/R-05/004aF-cF. Research Triangle Park, NC.


170 Smith, G., Coulston, J., Jepsen, E., Prichard, T. (2003). A national ozone biomonitoring program—results
from field surveys of ozone sensitive plants in Northeastern forests (1994-2000). Environmental Monitoring and
Assessment, 87, 271-291.

171 U.S. EPA (United States Environmental Protection Agency). 2006. Air quality criteria for ozone andrelated
photochemical oxidants. EPA/600/R-05/004aF-cF. Research Triangle Park, NC.


172US EPA. (2007) .Review of the National Ambient Air Quality Standards for Ozone: Policy assessment of
scientific and technical information. Office of Air Quality Planning and Standards staff paper. EPA-452/R-07-
003.

173 Chappelka, A.H., Samuelson, L.J. (1998). Ambient ozone effects on forest trees of the eastern United States:
a review. New Phytologist, 139,91-108.

174 U.S. EPA (2004) Air Quality Criteria for Particulate Matter (Oct 2004), Volume I Document No. EPA600/P-
99/002aF and Volume II Document No. EPA600/P-99/002bF.  This document is available in Docket EPA-HQ-
OAR-2005-0161.

175 U.S. EPA (2005) Review of the National Ambient Air Quality Standard for Particulate Matter: Policy
Assessment of Scientific and Technical Information, OAQPS Staff Paper.  EPA-452/R-05-005. This document
is available in Docket EPA-HQ-OAR-2005-0161.

176 U.S. EPA, 2008. Integrated Science Assessment for Oxides of Nitrogen and Sulfur- Ecological Criteria
(Final). U.S. EPA, Washington D.C., EPA/600/R-08/082F.

177 Environmental Protection Agency (2003). Response Of Surface Water Chemistry to the Clean Air Act
Amendments of 1990. National Health and Environmental Effects Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA 620/R-03/001.

178 Fenn, M.E. and Blubaugh, T.J. (2005) Winter Deposition of Nitrogen and Sulfur in the Eastern Columbia
River Gorge National Scenic Area, USDA Forest Service.

179 Galloway, J. N.; Cowling, E. B. (2002). Reactive nitrogen and the world: 200 years of change. Ambio 31: 64-
71.

180 Bricker, Suzanne B., et al., National Estuarine Eutrophication Assessment, Effects of Nutrient  Enrichment in
the Nation's Estuaries, National Ocean Service, National Oceanic  and Atmospheric Administration, September,
1999.
                                               7-72

-------
                                                           Environmental and Health Impacts
181 Smith, W.H. 1991. "Air pollution and Forest Damage." Chemical Engineering News, 69(45): 30-43.

182Gawel, J.E.; Ahner, B.A.; Friedland, A.J.; and Morel, F.M.M. 1996. "Role for heavy metals in forest decline
indicated by phytochelatin measurements." Nature, 381: 64-65.

183 Cotrufo,  M.F.; DeSanto, A.V.; Alfani, A.; et al. 1995. "Effects of urban heavy metal pollution on organic
matter decomposition in Quercus ilix L. woods." Environmental Pollution, 89: 81-87.

184 Niklinska, M.; Laskowski, R.; Maryanski, M. 1998. "Effect of heavy metals and storage time on two types of
forest litter: basal respiration rate and exchangeable metals." Ecotoxicological Environmental Safety, 41: 8-18.

185 Mason, R.P. and Sullivan, KA. 1997. "Mercury in Lake Michigan." Environmental Science & Technology,
31: 942-947. (from Delta Report "Atmospheric deposition of toxics to the Great Lakes").

186 Landis, M.S. and Keeler, G.J. 2002. "Atmospheric mercury deposition to Lake Michigan during the Lake
Michigan Mass Balance Study." Environmental Science & Technology, 21:  4518-24.

187U.S. EPA. 2000. EPA453/R-00-005, "Deposition of Air Pollutants to the  Great Waters: Third Report to
Congress," Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina.  Error! Main
Document Only.This document is available in Docket EPA-HQ-OAR-2004-0008.

188NSTC 1999

189 Callender, E. and Rice, K.C. 2000. "The Urban Environmental Gradient:  Anthropogenic Influences on the
Spatial and  Temporal Distributions of Lead and Zinc in Sediments." Environmental Science & Technology, 34:
232-238.

190 Rice, K.C. 1999. "Trace Element Concentrations in Streambed Sediment  Across the Conterminous United
States." Environmental Science & Technology, 33: 2499-2504.

191 Ely, JC; Neal, CR; Kulpa, CF; et al. 2001. "Implications of Platinum-Group Element Accumulation along
U.S.  Roads  from Catalytic-Converter Attrition." Environ. Sci. Technol. 35: 3816-3822.

192U.S. EPA. 1998. EPA454/R-98-014, "Locating and Estimating Air Emissions from Sources of Polycyclic
Organic Matter,"  Office of Air Quality Planning and  Standards, Research Triangle Park, North Carolina. This
document is available in Docket EPA-HQ-OAR-2004-0008.

193U.S. EPA. 1998. EPA454/R-98-014, "Locating and Estimating Air Emissions from Sources of Polycyclic
Organic Matter,"  Office of Air Quality Planning and  Standards, Research Triangle Park, North Carolina. Error!
Main Document Only.This document is available in Docket EPA-HQ-OAR-2004-0008.

194 Simcik, M.F.; Eisenreich, S.J.; Golden, K.A.; et al. 1996. "Atmospheric Loading of Polycyclic Aromatic
Hydrocarbons to Lake Michigan as Recorded in the Sediments." Environmental Science and Technology, 30:
3039-3046

195 Simcik, M.F.; Eisenreich, S.J.; and Lioy, P.J.  1999. "Source apportionment and source/sink relationship of
PAHs in the coastal atmosphere of Chicago and Lake Michigan." Atmospheric Environment, 33: 5071-5079.

196 Arzayus, K.M.; Dickhut, R.M.; and Canuel, E.A. 2001. "Fate of Atmospherically Deposited Polycyclic
Aromatic Hydrocarbons (PAHs) in Chesapeake Bay." Environmental Science & Technology, 35, 2178-2183.

197Park, J.S.; Wade, T.L.; and Sweet, S. 2001. "Atmospheric distribution of polycyclic aromatic hydrocarbons
and deposition to Galveston Bay, Texas, USA." Atmospheric Environment,  35: 3241-3249.
                                               7-73

-------
Draft Regulatory Impact Analysis
198 Poor, N.; Tremblay, R.; Kay, H.; et al. 2002. "Atmospheric concentrations and dry deposition rates of
polycyclic aromatic hydrocarbons (PAHs) for Tampa Bay, Florida, USA." Atmospheric Environment 38: 6005-
6015.

199 Arzayus, K.M.; Dickhut, R.M.; and Canuel, E.A. 2001. "Fate of Atmospherically Deposited Polycyclic
Aromatic Hydrocarbons (PAHs) in Chesapeake Bay." Environmental Science & Technology, 35, 2178-2183.

200U.S. EPA. 2000. EPA453/R-00-005, "Deposition of Air Pollutants to the Great Waters: Third Report to
Congress," Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina. This
document is available in Docket EPA-HQ-OAR-2004-0008.

201 Van Metre, P.C.; Mahler, B.J.; and Furlong, E.T. 2000. "Urban Sprawl Leaves its PAH Signature."
Environmental Science & Technology, 34: 4064-4070.

202 Cousins, I.T.; Beck, A.J.; and Jones, K.C. 1999. "A review of the processes involved in the exchange of semi-
volatile organic compounds across the air-soil interface." The Science of the Total Environment, 228: 5-24.

203 Tuhackova, J. et al. (2001) Hydrocarbon deposition and soil microflora as affected by highway traffic.
Environmental Pollution, 113: 255-262.

204US EPA. 1991. Effects of organic chemicals in the atmosphere on terrestrial plants. EPA/600/3-91/001.

205 cape JN, ID Leith, J Binnie,  J Content, M Donkin, M Skewes, DN Price AR Brown, AD Sharpe. 2003.
Effects of VOCs on herbaceous plants in an open-top chamber experiment. Environ. Pollut. 124:341-343.

206 Cape IN, ID Leith, J Binnie, J Content, M Donkin, M Skewes, DN Price AR Brown, AD Sharpe. 2003.
Effects of VOCs on herbaceous plants in an open-top chamber experiment. Environ. Pollut. 124:341-343.

207 Viskari E-L. 2000.  Epicuticular wax of Norway spruce needles as indicator of traffic pollutant deposition.
Water, Air, and Soil Pollut. 121:327-337.

208 Ugrekhelidze D, F Korte, G  Kvesitadze.  1997. Uptake and transformation of benzene and toluene by plant
leaves. Ecotox. Environ. Safety 37:24-29.

209 Kammerbauer H, H Selinger, R Rommelt, A Ziegler-Jons, D Knoppik, B Hock. 1987. Toxic components of
motor vehicle emissions for the spruce Pcieaabies. Environ. Pollut. 48:235-243.

210U.S. EPA. (2007). PM2.5 National Ambient Air Quality Standard Implementation Rule (Final). Washington,
DC: U.S. EPA. Retrieved on May 14, 2009  from Docket EPA-HQ-OAR-2003-0062 at
http://www.regulations.gOV/.72 FR 20586.
U.S. EPA. (2006J. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-
0190 at http://www.regulations.gov/.

211PM Standards Revision - 2006: Timeline. Retrieved on March 19, 2009 from
http://www.epa.gov/oar/particlepollution/naaqsrev2006.htmltftimeline
212 US EPA: 8-hour Ozone Nonattainment Areas. Retrieved on March 19, 2009
http: //www. epa.gov/ttn/atw/nata2002/ri sksum. html

213 Carbon Monoxide Nonattainment Area Summary:  http://www.epa.gov/air/oaqps/greenbk/cnsum.html

214 U.S. EPA. (2009) 2002 National-Scale Air Toxics Assessment.
http: //www. epa.gov/ttn/atw/natal 999/ri sksum. html

215 Control of Hazardous Air Pollutants From Mobile Sources (72 FR 8428; February 26, 2007)
                                               7-74

-------
                                                            Environmental and Health Impacts
216 US EPA (2007) Control of Hazardous Air Pollutants from Mobile Sources Regulatory Impact Analysis. EPA
document number 420-R-07-002, February 2007.

217U.S. Environmental Protection Agency, Byun, D.W., and Ching, J.K.S., Eds, 1999. Science algorithms of
EPA Models-3 Community Multiscale Air Quality (CMAQ modeling system, EPA/600/R-99/030, Office of
Research and Development).
218 Byun, D.W., and Schere, K.L., 2006. Review of the Governing Equations, Computational Algorithms, and
Other Components of the Models-3 Community Multiscale Air Quality (CMAQ) Modeling System, J. Applied
Mechanics Reviews, 59 (2), 51-77.
219 Dennis, R.L., Byun, D.W., Novak, J.H., Galluppi, K.J., Coats, C.J., and Vouk, M.A., 1996. The next
generation of integrated air quality modeling: EPA's Models-3, Atmospheric Environment, 30, 1925-1938.
220 US EPA (2007). Regulatory Impact Analysis of the Proposed Revisions to the National Ambient Air Quality
Standards for Ground-Level Ozone.  EPA document number 442/R-07-008, July 2007.
221 Aiyyer, A, Cohan, D., Russell, A., Stockwell, W, Tanrikulu, S., Vizuete, W., Wilczak, J, 2007. Final Report:
Third Peer Review of the CMAQ Model, p. 23.

222Grell, G., Dudhia, J., Stauffer, D. (1994). A Description of the Fifth-Generation Perm State/NCAR Mesoscale
Model (MM5), NCAR/TN-398+STR.,  138 pp, National Center for Atmospheric Research, Boulder CO

223Grell,  G., Dudhia, J., Stauffer, D. (1994). A Description of the Fifth-Generation Perm State/NCAR Mesoscale
Model (MM5), NCAR/TN-398+STR.,  138 pp, National Center for Atmospheric Research, Boulder CO.

224 Yantosca, B. (2004). GEOS-CHEMv7-01-02 User's Guide, Atmospheric Chemistry Modeling Group,
Harvard University, Cambridge, MA, October  15, 2004.

225 U.S. Environmental Protection Agency.  (2008). Final Ozone NAAQS Regulatory Impact Analysis.
Prepared by: Office of Air and Radiation, Office of Air Quality Planning and Standards. March.

226 U.S. Environmental Protection Agency.  October 2006. Final Regulatory Impact Analysis (RIA) for the
Proposed National Ambient Air Quality Standards for Particulate Matter.  Prepared by: Office of Air and
Radiation.

227 U.S. Environmental Protection Agency (U.S. EPA). 2009a. Regulatory Impact Analysis: National Emission
Standards for Hazardous Air Pollutants from the Portland Cement Manufacturing Industry. Office of Air
Quality Planning and Standards, Research Triangle Park, NC.  April. Available on the Internet at
.

228 U.S. Environmental Protection Agency (U.S. EPA). 2009b. Proposed NO2 NAAQS Regulatory Impact
Analysis (RIA).  Office of Air Quality Planning and Standards, Research Triangle Park, NC. April. Available
on the Internet at http://www.epa.gov/ttn/ecas/regdata/RIAs/proposedno2ria.pdf.

229 U.S. Environmental Protection Agency (U.S. EPA). 2008a. Regulatory Impact Analysis, 2008 National
Ambient Air Quality Standards for Ground-level Ozone, Chapter 6.  Office of Air Quality Planning and
Standards, Research Triangle Park, NC. March. Available at .

230 U.S. Environmental Protection Agency (U.S. EPA). 2009a. Regulatory Impact Analysis: National Emission
Standards for Hazardous Air Pollutants from the Portland Cement Manufacturing Industry. Office of Air
Quality Planning and Standards, Research Triangle Park, NC.  April. Available on the Internet at
.

231 U.S. Environmental Protection Agency (U.S. EPA). 2009b. Proposed NO2 NAAQS Regulatory Impact
Analysis (RIA).  Office of Air Quality Planning and Standards, Research Triangle Park, NC. April. Available
on the Internet at http://www.epa.gov/ttn/ecas/regdata/RIAs/proposedno2ria.pdf
                                               7-75

-------
Draft Regulatory Impact Analysis
232 U.S. Environmental Protection Agency (U.S. EPA).  2008b. Technical Support Document: Calculating
Benefit Per-Ton estimates, Ozone NAAQS Docket #EPA-HQ-OAR-2007-0225-0284. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. March. Available on the Internet at
.

233 Fann, N. et al. (2009). The influence of location, source, and emission type in estimates of the human health
benefits of reducing a ton of air pollution. Air Qual Atmos Health. Published online: 09 June, 2009.

234 Pope, C.A., III, R.T. Burnett, M.J. Thun, E.E. Calle,  D. Krewski, K. Ito, and G.D. Thurston.  2002. "Lung
Cancer,  Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution."  Journal of the
American  Medical Association 287:1132-1141.

235 Laden,  F., J. Schwartz, F.E. Speizer, and D.W. Dockery. 2006. "Reduction in Fine Particulate Air Pollution
and Mortality."  American Journal of Respiratory and Critical Care Medicine 173:667-672.  Estimating the
Public Health Benefits of Proposed Air Pollution Regulations. Washington, DC: The National Academies Press.

236 Roman, Henry A., Katherine D.  Walker, Tyra L. Walsh, Lisa Conner, Harvey M. Richmond, Bryan J.
Hubbell, and Patrick L. Kinney. 2008. Expert Judgment Assessment of the Mortality Impact of Changes in
Ambient Fine Particulate Matter in  the U.S. Environ. Sci. Technol., 42(7):2268-2274.

237 Industrial Economics, Inc. 2006. Expanded Expert Judgment Assessment of the Concentration-Response
Relationship Between PM2.5 Exposure and Mortality. Prepared for the U.S. EPA, Office of Air Quality
Planning and Standards, September. Available on the Internet at


238 Mrozek, J.R., and L.O. Taylor. 2002. "What Determines the Value of Life? A Meta-Analysis." Journal of
Policy Analysis and Management 21(2):253-270.

239 Viscusi, V.K., and J.E. Aldy. 2003. "The Value of a Statistical Life: A Critical Review of Market Estimates
throughout the World."  Journal of  Risk and Uncertainty 27(1): 5-76.

240 Kochi,  L, B. Hubbell, and R. Kramer. 2006. An Empirical Bayes Approach to Combining Estimates of the
Value of Statistical Life for Environmental Policy Analysis. Environmental and Resource Economics.  34: 385-
406.

241 U.S. Environmental Protection Agency (U.S. EPA).  2000. Guidelines for Preparing Economic Analyses.
EPA 240-R-00-003. National Center for Environmental Economics, Office of Policy Economics and
Innovation. Washington, DC.  September.  Available on the Internet at
.

242 Information on BenMAP, including downloads of the software, can be found at http://www.epa.gov/ttn/ecas/
benmodels.html.

243 Bell,  M.L., et al. (2004). Ozone and short-term mortality in 95 US urban communities, 1987-2000. JAMA,
2004. 292(19):  p. 2372-8.

244 Huang, Y.; Dominici, F.; Bell, M. L. (2005) Bayesian hierarchical distributed lag models for summer ozone
exposure and cardio-respiratory mortality. Environmetrics. 16: 547-562.

245 Schwartz, J. (2005) How sensitive is the association between ozone and daily deaths to control for
temperature?Am. J. Respir. Crit. CareMed. 111'. 627-631.

246 Bell,  M.L., F. Dominici, and J.M. Samet. (2005). A meta-analysis of time-series studies of ozone and
mortality with comparison to the national morbidity, mortality, and air pollution study. Epidemiology. 16(4): p.
436-45.
                                                7-76

-------
                                                            Environmental and Health Impacts
247 Ito, K., S.F. De Leon, and M. Lippmann (2005). Associations between ozone and daily mortality: analysis
and meta-analysis. Epidemiology. 16(4): p. 446-57.

248 Levy, J.I., S.M. Chemerynski, and J.A. Sarnat. (2005). Ozone exposure and mortality: an empiric bayes
metaregression analysis. Epidemiology. 16(4): p. 458-68.

249 Pope, C.A., III, R.T. Burnett, M.J. Thun, E.E. Calle, D. Krewski, K. Ito, and G.D. Thurston. (2002). "Lung
Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution."  Journal of the
American Medical Association 287:1132-1141.

250 Laden, F.,  J. Schwartz, F.E.  Speizer, and D. W. Dockery.  (2006). Reduction in Fine Particulate Air
Pollution and Mortality.  American Journal of Respiratory and Critical Care Medicine.  173:  667-672.

251 Industrial Economics, Incorporated (lEc). (2006). Expanded Expert Judgment Assessment of the
Concentration-Response Relationship Between PM2.5 Exposure and Mortality.  Peer Review Draft.  Prepared
for: Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, NC. August.

252 Woodruff, T.J., J. Grille, and K.C. Schoendorf (1997). The Relationship Between Selected Causes of
Postneonatal Infant Mortality and Particulate Air Pollution in the United States.  Environmental Health
Perspectives. 105(6):608-612.

253 Abbey, D.E., B.L. Hwang, R.J. Burchette, T. Vancuren, and P.K.  Mills. (1995). Estimated Long-Term
Ambient Concentrations of PM(10) and Development of Respiratory Symptoms in a Nonsmoking Population.
Archives of Environmental Health. 50(2): 139-152.

254 Peters, A., D.W. Dockery, J.E. Muller, and M.A. Mittleman. (2001). Increased Particulate Air Pollution and
the Triggering  of Myocardial Infarction. Circulation. 103:2810-2815.

255 Schwartz J. (1995). Short term fluctuations in air pollution and hospital admissions of the elderly for
respiratory disease. Thorax. 50(5):531-538.

256 Schwartz J. (1994a). PM(10) Ozone, and Hospital Admissions For the Elderly in Minneapolis St Paul,
Minnesota. Arch Environ Health. 49(5):366-374.

257 Schwartz J. (1994b). Air Pollution and Hospital Admissions For the Elderly in Detroit, Michigan.  Am J
RespirCritCareMed. 150(3):648-655.

258 Moolgavkar SH, Luebeck EG, Anderson EL. (1997). Air pollution and hospital admissions for respiratory
causes in Minneapolis St. Paul and Birmingham.  Epidemiology. 8(4):364-370.

259 Burnett RT, Smith-Doiron M, Stieb D, Raizenne ME, Brook JR, Dales RE, et al. (2001). Association
between ozone and hospitalization for acute respiratory diseases in children less than 2 years of age.  Am J
Epidemiol. 153(5):444-452.

260 Moolgavkar, S.H.  (2003). "Air Pollution and Daily Deaths and Hospital  Admissions in Los Angeles and
Cook Counties." In Revised Analyses of Time-Series Studies of Air Pollution andHealth. Special Report.
Boston, MA:  Health Effects Institute.

261 Ito, K. (2003). "Associations of Particulate Matter Components with Daily Mortality and Morbidity in
Detroit, Michigan." In Revised Analyses of Time-Series Studies of Air Pollution andHealth. Special Report.
Health Effects  Institute, Boston, MA.

262 Moolgavkar, S.H.  (2000). Air Pollution and Hospital Admissions for Diseases of the Circulatory System in
Three U.S. Metropolitan Areas. Journal of the Air and Waste Management Association 50:1199-1206.

263 Sheppard, L.  (2003). Ambient Air Pollution and Nonelderly Asthma Hospital Admissions in Seattle,
Washington, 1987-1994. InRevised Analyses of Time-Series Studies of Air Pollution andHealth.  Special
Report.  Boston,  MA:  Health Effects Institute.

264 Jaffe DH, Singer ME, Rimm AA.  (2003).  Air pollution and emergency department visits for asthma among
Ohio Medicaid recipients, 1991-1996. Environ Res 91(l):21-28.


                                                7-77

-------
Draft Regulatory Impact Analysis
265 Peel, J. L., P. E. Tolbert, M. Klein, et al. (2005). Ambient air pollution and respiratory emergency department
visits. Epidemiology. Vol. 16 (2): 164-74.

266 Wilson, A. M., C. P. Wake, T. Kelly, et al. (2005). Air pollution, weather, and respiratory emergency room
visits in two northern New England cities: an ecological time-series study. Environ Res. Vol. 97 (3): 312-21.

267Norris, G., S.N. YoungPong, J.Q. Koenig, T.V. Larson, L. Sheppard, and J.W. Stout.  (1999). An
Association between Fine Particles and Asthma Emergency  Department Visits for Children in Seattle.
Environmental Health Perspectives 107(6):489-493.

268 Dockery, D.W., J. Cunningham, A.I. Damokosh, L.M. Neas, J.D. Spengler, P. Koutrakis, J.H. Ware, M.
Raizenne, and F.E. Speizer.  (1996).  Health Effects of Acid Aerosols On North American Children-Respiratory
Symptoms. Environmental Health Perspectives 104(5): 500-505.

269 Pope, C.A., III, D.W. Dockery, J.D. Spengler, and M.E. Raizenne. (1991). Respiratory Health and PM10
Pollution: A Daily Time Series Analysis. American Review of Respiratory Diseases 144:668-674.

270 Schwartz, J., and L.M. Neas. (2000). Fine Particles are More Strongly Associated than Coarse Particles with
Acute Respiratory Health Effects in Schoolchildren. Epidemiology 11:6-10.

271 Ostro, B., M. Lipsett, J. Mann, H. Braxton-Owens, and M. White. (2001). Air Pollution and Exacerbation of
Asthma in African-American Children in Los Angeles. Epidemiology 12(2):200-208.

272 Vedal, S., J. Petkau, R. White, and J. Blair.  (1998). Acute Effects of Ambient Inhalable Particles in
Asthmatic and Nonasthmatic Children.  American Journal of Respiratory and Critical Care Medicine
157(4):1034-1043.

273 Ostro, B.D. (1987). Air Pollution and Morbidity Revisited: A Specification Test. Journal of Environmental
Economics Management 14:87-98.

274 Gilliland FD, Berhane K, Rappaport EB,  Thomas DC, Avol E, Gauderman WJ, et al. (2001). The effects of
ambient air pollution on school absenteeism due to respiratory illnesses. Epidemiology 12(l):43-54.

275 Chen L, Jennison BL, Yang W, Omaye ST. (2000).  Elementary school absenteeism and air pollution. Inhal
laaco/12(ll):997-1016.

276 Ostro, B.D. and S. Rothschild.  (1989). Air Pollution and Acute Respiratory Morbidity: An Observational
Study of Multiple Pollutants. Environmental Research 50:238-247.

277 Russell, M.W., D.M. Huse, S. Drowns, B.C. Hamel, and  S.C. Hartz.  (1998).  Direct Medical Costs of
Coronary Artery Disease in the United States. American Journal of Cardiology 81(9): 1110-1115.

278 Wittels, E.H, J.W. Hay, and A.M. Gotto, Jr. (1990).  Medical Costs of Coronary Artery Disease in the
United States. American Journal of Cardiology 65(7):432-440.

279 Smith, D.H., D.C. Malone, K.A. Lawson, L.J. Okamoto,  C. Battista, and W.B. Saunders. (1997).  A National
Estimate of the Economic Costs of Asthma.  American Journal of Respiratory and Critical Care Medicine 156(3
Ptl):787-793.

280 Stanford, R., T. McLaughlin, and L.J. Okamoto. (1999).  The Cost of Asthma in the Emergency Department
and Hospital. American Journal of Respiratory and Critical Care Medicine 160(1):211-215.

281 Rowe, R.D., and L.G. Chestnut. (1986).  Oxidants and Asthmatics in Los Angeles: A Benefits Analysis—
Executive Summary. Prepared by Energy and Resource Consultants, Inc.  Report to the U.S. Environmental
Protection Agency, Office of Policy Analysis.  EPA-230-09-86-018. Washington, DC.

282 Science Advisory Board. 2001. NATA — Evaluating the National-Scale Air Toxics Assessment for 1996 —
an SAB Advisory, http://www.epa.gov/ttn/atw/sab/sabrev.html.

283 MiniCAM is a long-term, global integrated assessment model of energy, economy, agriculture and land use,
that considers the sources of emissions of a suite of greenhouse gases (GHG's), emitted in 14 globally
disaggregated regions, the fate of emissions to the atmosphere, and the consequences of changing concentrations


                                                7-78

-------
                                                            Environmental and Health Impacts
of greenhouse related gases for climate change. MiniCAM begins with a representation of demographic and
economic developments in each region and combines these with assumptions about technology development to
describe an internally consistent representation of energy, agriculture, land-use, and economic developments that
in turn shape global emissions.

Brenkert A, S. Smith, S.  Kim,  and H. Pitcher, 2003:  Model Documentation  for the  MiniCAM. PNNL-14337,
    Pacific Northwest National Laboratory, Richland, Washington.

284 MAGICC consists of a suite of coupled gas-cycle, climate and ice-melt models integrated into a single
framework. The framework allows the user to determine changes in greenhouse-gas concentrations, global-mean
surface air temperature and sea-level resulting from anthropogenic emissions of carbon dioxide (CO2), methane
(CH4), nitrous oxide (N2O), reactive gases (CO, NOx, VOCs), the halocarbons (e.g. HCFCs, HFCs, PFCs) and
sulfur dioxide (SO2). MAGICC emulates the global-mean temperature responses of more sophisticated coupled
Atmosphere/Ocean General Circulation Models (AOGCMs) with high accuracy.

Wigley, T.M.L. and Raper,  S.C.B. 1992. Implications for Climate And Sea-Level of Revised IPCC Emissions
    Scenarios Nature 357, 293-300. Raper, S.C.B., Wigley T.M.L. and Warrick R.A. 1996. in Sea-Level Rise
    and  Coastal Subsidence: Causes, Consequences  and Strategies J.D.  Milliman, B.U. Haq, Eds., Kluwer
    Academic Publishers, Dordrecht, The Netherlands, pp. 11-45.
Wigley, T.M.L. and  Raper, S.C.B. 2002. Reasons for larger warming projections in the IPCC Third Assessment
    Report J. Climate 15, 2945-2952.

285This scenario is used because it contains a comprehensive suite of greenhouse and pollutant gas emissions.
The four RCP scenarios will be used as common inputs into a variety of Earth System Models for inter-model
comparisons leading to the IPCC AR5 (Moss et al. 2008). The MiniCAM RCP4.5 is based on the scenarios
presented in Clarke et al. (2007) with non-CO2 and pollutant gas emissions implemented as described in Smith
and Wigley (2006). Base-year information has been updated to the latest available data for the RCP process. The
final RCP4.5 scenario will be available at the IAMC scenario Web site (www.iiasa.ac.at/web-apps/tnt/RcpDb/).

Clarke, L., J. Edmonds,  H. Jacoby, H. Pitcher, J. Reilly, R.  Richels, (2007) Scenarios of  Greenhouse Gas
    Emissions and Atmospheric Concentrations. Sub-report 2.1 A of Synthesis and Assessment Product 2.1 by
    the U.S. Climate Change Science Program and the Subcommittee on Global Change Research  (Department
    of Energy, Office of Biological & Environmental Research, Washington, DC., USA, 154 pp.).

Moss, Richard, Mustafa Babiker, Sander Brinkman, Eduardo Calvo, Tim Carter, Jae Edmonds, Ismail Elgizouli,
    Seita Emori,  Lin Erda, Kathy Hibbard, Roger Jones, Mikiko Kainuma, Jessica Kelleher, Jean Francois
    Lamarque, Martin Manning, Ben Matthews, Jerry Meehl, Leo Meyer, John  Mitchell, Neboj sa Nakicenovic,
    Brian O'Neill, Ramon Pichs, Keywan Riahi, Steven Rose, Paul Runci, Ron Stouffer, Detlef  van Vuuren,
    John Weyant, Tom Wilbanks, Jean Pascal van Ypersele, and Monika Zurek (2008) Towards New Scenarios
    for Analysis of Emissions, Climate Change, Impacts, and Response Strategies (Intergovernmental Panel on
    Climate Change, Geneva) 132 pp.

Smith, Steven J.  and T.M.L.  Wigley (2006)  "Multi-Gas Forcing Stabilization with the  MiniCAM" Energy
    Journal (Special Issue #3).

286 In IPCC reports, equilibrium climate sensitivity refers to the equilibrium change in the annual mean global
surface temperature following a doubling of the atmospheric equivalent carbon dioxide concentration. The IPCC
states that climate sensitivity is "likely" to be in the range of 2°C to 4.5°C, "very unlikely" to be less than 1.5°C,
and "values substantially higher than 4.5°C cannot be excluded." IPCC WGI, 2007, Climate Change 2007 - The
Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the IPCC,
http://www.ipcc.ch/.

287 IPCC WGI, 2007. The baseline temperature increases by 2100 from our MiniCAM-MAGICC runs are 1.8°C
to4.5°C.
                                               7-79

-------
Draft Regulatory Impact Analysis
288 IPCC WGII. 2007. Climate Change 2007 - Impacts, Adaptation and Vulnerability Contribution of Working
Group II to the Fourth Assessment Report of the IPCC.

289 "Light-Duty Automotive Technology and Fuel Economy Trends: 1975 Through 2008", EPA420-R-08-015,
U.S. Environmental Protection Agency Office of Transportation and Air Quality, September 2008

290 "Future Generation Passenger Compartment-Validation" in "Lightweighting Materials - FY 2008 Progress
Report", U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Vehicle Technologies
Program, May 2009.

291 "Relationship between Vehicle Size and Fatality Risk in Model Year 1985-93 Passenger cars and Light
Trucks", DOT HS 808 570, NHTSA Technical Report, January 1997

292 "Vehicle Weight, Fatality Risk and  Crash Compatibility of Model Year 1991-99 Passenger Cars and Light
Trucks", DOT HS 809 663, NHTSA Technical Report, October 2003

293 "Supplemental Results on the Independent Effects of Curb Weight, Wheelbase and Track on Fatality Risk",
Dynamic Research,  Inc., DRI-TR-05-01, May 2005

294 "An Assessment of the Effects of Vehicle Weight and Size on Fatality Risk in 1985 to 1998 Model Year
Passenger Cars and  1985 to  1997 Model Year Light Trucks and Vans",  Van Auken, M., Zellner J.W., SAE
Technical Paper Number 2005-01-1354, 2005.

295 FR Vol. 74, No. 59, beginning on pg. 14402
                                              7-80

-------
                                                  Other Economic and Social Impacts

CHAPTER 8: Other Economic and Social Impacts

8.1 Vehicle Sales Impacts

8.1.1 How Vehicle Sales Impacts were Estimated for this Rule

       The vehicle sales impacts discussed in Section III.H.4 of the preamble to the proposal
and presented below in Table 8-1 and Table 8-2 were derived using the following
methodology. For additional discussion of the assumptions used in the vehicles sales impacts,
see Section III.H of the preamble. The calculation is performed for an average car and an
average truck, rather than for individual vehicles. The analysis conducted for this rule does
not have the precision to examine effects on individual manufacturers or different vehicle
classes. Chapter 8.1.2 provides our assessment of models that examine these questions.

       The analysis starts with the increase in costs estimated by OMEGA.  We assume that
these costs are fully passed along to consumers. This assumption is appropriate for cost
increases in perfectly competitive markets.  In less than perfectly competitive markets,
though, it is likely that the cost increase is split between consumers and automakers, and the
price is not likely to increase as much as costs.1 Thus, the assumption of full cost pass-
through is probably an overestimate,  and price is not likely to increase as much as estimated
here.

       The next step in the analysis is to adjust this cost increase for other effects on the
consumer.  We assume that the consumer holds onto this vehicle  for 5 years and then sells it.
The higher vehicle price  is likely to lead to an increase in sales tax, insurance, and vehicle
financing costs, as well as increases in the resale value of the vehicle.  These factors weigh
against each other: the higher sales tax and insurance costs increase costs to consumers; the
higher resale value allows consumers to recover a portion of these costs.

       The increase in insurance costs is estimated from the average value of collision plus
comprehensive insurance as a proportion of average new vehicle  price. Collision plus
comprehensive insurance is the portion of insurance costs that depend on vehicle value. The
Insurance Information Institute2 provides the average value of collision plus comprehensive
insurance in 2006  as $448. The average value of a new vehicle in 2006, according to the U.S.
Department of Energy, was $22,651.3 (This value is for a 2006 vehicle in 2006 and is used
only for the insurance adjustment; it does not correspond to the new vehicle prices, described
below, used in the vehicle sales impact calculation.)  Dividing the insurance cost by the
average price of a new vehicle gives the  proportion of comprehensive plus collision insurance
as 1.98% of the price of a vehicle.  If this same proportion holds for the increase in price of a
vehicle, then insurance costs should go up by 1.98% of the increase in vehicle cost.  For the
five-year period, the present value of this increase in insurance cost would be worth 9.0% of
the vehicle cost increase, using a 3% discount rate (8.1% at a  7% discount rate).

       Calculating the average increase in sales tax starts with the vehicle sales tax for each
state in 2006.4  The sales tax per state was then multiplied by the  2006 population of the
state;5 those values were summed and divided by total U.S.  population, to give a population-
                                         8-1

-------
Draft Regulatory Impact Analysis

weighted sales tax. That estimate of the state sales taxes for vehicles in the U.S. is 5.3% in
2006. This value is assumed to be a one-time cost incurred when the vehicle is purchased.

       As of August 24, 2009, the national average interest rate for a 5 year new car loan was
7.41 percent.6  Converting the up-front payment to an annual value paid over five years results
in a consumer paying 24.7% of the up-front amount every year. The present value of these
five payments results in an increase of 12.9% of the cost, using a 3% discount rate; with a 7%
discount rate, the increase is 1.1%.  NHTSA's PRIA notes that 70% of auto purchases use
financing; applying that fraction to this cost increase results in an addition of 9.0% in
financing costs with a 3% discount rate, and 0.8% for a 7% discount rate.

       The average resale price of a vehicle after 5 years is about 35%7 of the original
purchase price.  Because the consumer can recover that amount after 5 years, it reduces the
effect of the increased cost of the vehicle.  Discounted to a present value at a 3% interest rate,
the increase in price should be worth about 30.2% to the vehicle purchaser (25.0% at a 7%
discount rate).  This approach is premised on the idea that the resale value of a vehicle is
directly proportional to the initial value, and that proportion does not change.

       Thus, the effect on a consumer's expenditure of the cost of the new technology (with
some rounding) should be (1 + 0.090 + 0.053 + 0.090 - 0.302) = 0.932 times the cost of the
technology at a 3% discount rate. At a 7% discount rate, the effect on a consumer's
expenditure of the cost of the new technology should be (1 + 0.081 + 0.053 + 0.008 - 0.250)
= 0.892 times the cost of the technology.

       The fuel cost savings are based on the five years of consumer ownership of the
vehicle. The analysis is done for each model-year for an average vehicle. Section 5.6 of this
DRIA discusses the source of aggregate fuel savings, in gallons, for cars and trucks for each
model year by year. These values  are divided by the total number of the vehicles produced to
get per-vehicle savings peryear for the first five years of the vehicle's life. This method
ignores the few vehicles of the new model year that are scrapped. Because incorporating
scrappage would reduce the denominator, and thus increase per-vehicle fuel savings, it
underestimates per-vehicle fuel savings by a small amount. The per-vehicle fuel savings in
gallons are multiplied by the price  of fuel to get the per-vehicle fuel savings in dollars.  For
each model year, then, the first five years of fuel savings are discounted and summed to
produce the present value of fuel savings for that vintage vehicle. For instance, the 2016 fuel
savings per vehicle are the present value in year 2016 of fuel savings estimated for 2016
through 2020.

       The prices for new vehicles are assumed to be constant at the 2008 value (in 2007$) of
$26,201 for a car, and $29,678 for a truck. These are the values used in NHTSA's 2011 rule
on CAFE standards.

       The fuel cost savings are subtracted from the increase in costs associated with the rule
to get the net effect of the rule on consumer expenditure. The higher cost leads consumers to
purchase fewer new vehicles, but the fuel savings can counteract this effect. This calculation
uses an elasticity of demand for new vehicles of -Is: that is, an increase of 1% in the price of
a new vehicle will lead to a 1% reduction in new vehicle sales. Using this value assumes that

                                         8-2

-------
                                                    Other Economic and Social Impacts
the demand elasticity for new vehicles under this rule is the same as the elasticity for older
vehicles. This change in consumer expenditure as a percent of the average price of a new
vehicle, with the elasticity of demand of -1, is the negative of the percent change in vehicle
purchases.  The net effect of this calculation on vehicle purchases is in Table 8-1 and Table
8-2.

                   Table 8-1 Vehicle Sales Impacts Using a 3% Discount Rate

2012
2013
2014
2015
2016
CHANGE IN
CAR SALES
66,600
93,300
134,400
236,300
375,400
% CHANGE
0.7
0.9
1.3
2.2
3.4
CHANGE IN
TRUCK SALES
27,300
161,300
254,400
368,400
519,000
% CHANGE
0.5
2.8
4.4
6.5
9.4
       Table 8-1 shows vehicle sales increasing.  Because the fuel savings associated with this
rule are expected to exceed the technology costs, the effective prices of vehicles - the
adjusted increase in technology cost less the fuel savings over five years ~ to consumers will
fall, and consumers will buy more new vehicles.  This effect is expected to increase over time.
As a result, if consumers consider fuel savings at the time that they make their vehicle
purchases, the lower net cost of the vehicles is expected to lead to an increase in sales for both
cars and trucks.  Both the absolute and the percent increases for truck sales are larger than
those for cars (except in 2012).

                   Table 8-2 Vehicle Sales Impacts Using a 7% Discount Rate

2012
2013
2014
2015
2016
CHANGE IN
CAR SALES
61,900
86,600
125,200
221,400
353,100
% CHANGE
0.7
0.9
1.2
2
3.2
CHANGE IN
TRUCK SALES
25,300
60,000
122,900
198,100
291,500
% CHANGE
0.5
1
2.1
3.5
5.3
       Table 8-2 shows the same calculations using a 7% discount rate. Qualitatively, the
results are identical to those using a 3% discount rate: the fuel savings outweigh the increase
in technology costs for all years. As a result, vehicle sales are expected to be higher under
this rule than in the absence of the rule. In addition, while the increased numbers of car sales
are larger than the  numbers for trucks, the percent increases are larger for trucks.

       This calculation focuses on changes in consumer expenditures as the explanatory
variable for changes in aggregate new vehicle sales. This is a simplification, since consumers
typically consider a number of factors in addition to expenditures when they decide on
purchasing a vehicle.  In addition, it does not consider changes in the mix of vehicles sold that

-------
Draft Regulatory Impact Analysis

may result from this rule. The next section discusses more complex modeling of the vehicle
purchase decision.

8.1.2 Consumer Vehicle Choice Modeling

       In this section we describe some of the consumer vehicle choice models EPA has
reviewed in the literature, and we describe the models' results and limitations that we have
identified.  The evidence from consumer vehicle choice models indicates a huge range of
estimates for consumers' willingness to pay for additional fuel economy. Because consumer
surplus estimates from consumer vehicle choice models depend critically on this value, we
would consider any consumer surplus estimates of the effect of our rule from such models to
be unreliable. In addition, the predictive ability of consumer vehicle choice models may be
limited. While vehicle choice models are based on sales of existing vehicles, vehicle models
are likely to change, both independently and in response to this proposed rule. The models
may not predict well in response to these changes. Instead, we compare the value of the fuel
savings associated with this rule with the increase in technology costs. Like NHTSA, EPA
will continue its efforts to review the literature, but, given the known difficulties, neither
NHTSA nor EPA has conducted an analysis using these models for this proposal.

       This rule will lead automakers to change characteristics - in particular, the fuel
economy ~ of the vehicles they produce.  These changes will affect the cost of manufacturing
the vehicle; as a result, the prices of the vehicles will also change.

       In response to these changes, the number and types of vehicles sold is likely to change.
When consumers buy vehicles, they consider both their personal characteristics (such as age,
family composition, income, and their vehicle needs) and the characteristics of vehicles (e.g.,
vehicle size, fuel economy, and price). In response to the changes in vehicle characteristics,
consumers  will reconsider their purchases. Increases in fuel economy are likely to be
attractive to consumers, but increases in price, as well as some changes in other vehicle
characteristics, may be deterrents to purchase. As a result, consumers may choose a different
vehicle than they would  have purchased in the absence of the rule.  The changes in prices and
vehicle characteristics are likely to influence consumers on multiple market scales:  the total
number of new vehicles  sold; the mix of new vehicles sold; and the effects  of the sales on the
used vehicle market.

       Consumer vehicle choice modeling (CCM) is a method used to predict what vehicles
consumers  will purchase, based on vehicle characteristics and prices. In principle, it should
produce more accurate estimates of compliance costs compared to models that hold fleet mix
constant, since it predicts changes in the fleet mix that can affect compliance costs.  It can also
be used to measure changes in consumer surplus, the benefit that consumers perceive from a
good over and above the purchase price. (Consumer surplus is the difference between what
consumers  would be willing to pay for a good, represented by the demand curve, and the
amount they actually pay. For instance, if a consumer were willing to pay $30,000 for a new
vehicle, but ended up paying $25,000, the $5000 difference is consumer surplus.)

       A number of consumer vehicle choice models have been developed. They vary in the
methods used, the data sources, the factors included in the models, the research questions they

                                         8-4

-------
                                                   Other Economic and Social Impacts

are designed to answer, and the results of the models related to the effects of fuel economy on
consumer decisions.  This section will give some background on these differences among the
models.

    8.1.2.1  Methods

       Consumer choice models (CCMs) of vehicle purchases typically use a form of discrete
choice modeling.  Discrete choice models seek to explain discrete rather than continuous
decisions. An example of a continuous decision is how many pounds of food a farm might
grow:  the pounds of food can take any numerical value. Discrete decisions can take only a
limited set of values. The decision to purchase a vehicle, for instance, can only take two
values, yes or no.  Vehicle purchases are typically modeled as discrete choices, where the
choice is whether to purchase a specified vehicle. The result of these models is a prediction
of the probability that a consumer will purchase a specified vehicle.  A minor variant on
discrete choice models estimates the market share for each vehicle.  Because the market share
is, essentially,  the probability that consumers will purchase a specific vehicle, these
approaches are similar in process; they differ mostly in the kinds of data that they use.

       The primary methods used to model vehicle choices are nested logit and mixed logit.
In a nested logit, the model is structured in  layers. For instance, the first layer may be the
choice of whether to buy a new or used vehicle. Given that the person chooses a new vehicle,
the second layer may be whether to buy a car or a truck;  given that the person chooses a car.
The third layer may be the choice among an economy, midsize, or luxury car. Examples of
nested logit models include  Goldberg,9 Greene et al.,10 and McManus.11

       In a mixed logit, personal characteristics of consumers play a larger role than in nested
logit.  While nested logit can look at the effects of a change in average consumer
characteristics, mixed logit allows consideration of the effects of the distribution of consumer
characteristics. As a result,  mixed logit can be used to examine the distributional effects on
various socioeconomic groups, which nested logit is not designed to do. Examples of mixed
logit models include Berry, Levinsohn, and Pakes,12 Bento et al.,13 and Train and Winston.14

       While discrete choice modeling appears to be the primary method for consumer choice
modeling, others (such as Kleit15 and Austin and Dinan16) have used a matrix of demand
elasticities to estimate the effects of changes in cost. The discrete choice models can produce
such elasticities. Kleit as well as Austin and Dinan used the elasticities from an internal GM
vehicle choice model.

    8.1.2.2  Data Sources

       The predictions of vehicle purchases from CCMs are  based on consumer and vehicle
characteristics. The CCMs identify the effects of changing the characteristics on the purchase
decisions. These effects are typically called the parameters or coefficients of the models. For
instance, the model parameters might predict that an increase in a person's income of 10%
would increase the probability of her purchasing vehicle A by 5%, and decrease the
probability of her purchasing vehicle B by  10%.
                                         8-5

-------
Draft Regulatory Impact Analysis

       The parameters in CCMs can be developed either from original data sources
(estimated models), or using values taken from other studies (calibrated models).

       Estimated models use datasets on consumer purchase patterns, consumer
characteristics, and vehicle characteristics to develop their original sets of parameters. The
datasets used in these studies sometimes come from surveys of individuals' behaviors.17
Because they draw on the behavior of individuals, they provide what is sometimes called
micro-level data. Other studies, that estimate market shares instead of discrete purchase
decisions, use aggregated data that can cover long time periods.18

       Calibrated models rely on existing studies for their parameters. Researchers may draw
on results from a number of estimated models, or even from research other than CCM, to
choose the parameters of the models.  The Fuel Economy Regulatory Analysis Model
developed for the Energy Information Administration19 and the New Vehicle Market Model
developed by NERA Economic Consulting20 are examples of calibrated models.

    8.1.2.3  Factors Included in the Models

       Consumer choice models vary in their complexity and levels of analysis. Some focus
only on the new vehicle market;21 others consider the choice between new vehicles and an
outside good (possibly including a used vehicle);22 others explicitly consider the relationship
between the new and used vehicle markets.23 Some models include consideration of vehicle
miles traveled,24 though most do not.

       The models vary  in their inclusion of both consumer and vehicle information.  One
model includes only vehicle price and the distribution of income  in the population influencing
choice;25 others include varying numbers and kinds of vehicle and consumer attributes.

    8.1.2.4  Research Questions for the Models

       Consumer choice models have been developed to analyze many different research and
policy questions. In part, these models have been developed to advance the state of economic
modeling. The work of Berry, Levinsohn, and Pakes,26 for instance, is often cited outside the
motor vehicle context for its incorporation of multiple new modeling issues  into its
framework.  In addition,  because the vehicle sector is a major part of the U.S. economy and a
stakeholder in many public policy discussions, research questions cover a wide gamut. These
topics have included the  effects of voluntary export restraints on  Japanese vehicles compared
to tariffs and quotas,27 the market acceptability of alternative-fuel vehicles,28 the effects of
introduction and exit of vehicles from markets,29 causes of the decline in market shares of
U.S. automakers,30 and the effects of gasoline taxes31 and "feebates"32 (subsidizing fuel-
efficient cars with revenue collected by taxing fuel-inefficient vehicles).

    8.1.2.5  The Effect of Fuel Economy on Consumer Decisions

       Consumer vehicle choice models typically consider the effect of fuel economy on
vehicle purchase decisions.  It can appear in various forms.

-------
                                                   Other Economic and Social Impacts

       Some models33 incorporate fuel economy through its effects on the cost of owning a
vehicle. With assumptions on the number of miles traveled per year and the cost of fuel, it is
possible to estimate the fuel savings (and perhaps other operating costs) associated with a
more fuel-efficient vehicle.  Those savings are considered to reduce the cost of owning a
vehicle: effectively, they reduce the purchase price. This approach relies on the assumption
that, when purchasing vehicles, consumers can estimate the fuel savings that they expect to
receive from a more fuel-efficient vehicle and consider the savings equivalent to a reduction
in purchase price. Turrentine and Kurani34 question this assumption; they find, in fact, that
consumers do not make this calculation when they purchase a vehicle. The question remains,
then, how or whether consumers take fuel economy into account when they purchase their
vehicles.

       Most estimated consumer choice models, instead of making assumptions about how
consumers incorporate fuel economy into their decisions, use data on consumer behavior to
identify that effect. In some models, the miles per gallon of vehicles is one of the vehicle
characteristics included to explain purchase decisions. Other models use fuel consumption
per mile, the inverse of miles per gallon, as a measure:35 since consumers pay for gallons of
fuel, then this measure can assess fuel savings  relatively directly.36  Yet other models multiply
fuel consumption per mile by the cost of fuel to get the price of driving a mile,37 or they
divide  fuel economy by fuel cost to get miles per dollar.38  It is worth noting that these last
two measures assume that consumers respond the same way to an increase in fuel economy as
they do to a  decrease in the price of fuel when  each has the same effect on cost per mile
driven. On the one hand, while this assumption does not rely on as complex a calculation as
the present value of fuel savings that Turrentine and Kurani examined, it suggests a
calculating consumer.  On the other hand, it is  also a way to recognize the role of fuel prices
in consumers' purchase of fuel economy: Busse et al.39 present results that higher fuel prices
play a  major role in that decision.

       Greene and Liu,40 in a paper published  in 1988, reviewed 10 papers using consumer
vehicle choice models and estimated for each one how much consumers would be willing to
pay at time of purchase to reduce vehicle  operating costs by $1 per year. They found that
people were willing to pay between $0.74 and  $25.97 for a $1 decrease in annual operating
costs for a vehicle. This is clearly a very  wide range: while the lowest estimate suggests that
people are not willing to pay $ 1 once to get $ 1  per year reduced costs of operating their
vehicles, the maximum suggests a willingness  to pay 35 times as high. For comparison, the
present value of saving $1 peryear for 15 years at a 3% discount rate is $11.94, while a 7%
discount rate produces a present value of  $8.78. While this study is quite old, it suggests that,
at least as of that time, consumer vehicle choice models produced widely varying estimates of
the value of reduced vehicle operating costs.

       More recent studies do not suggest agreement on the value of increased fuel economy
to consumers. For instance, some papers41 find that the role of fuel cost (price per gallon
divided by miles per gallon, or the cost of driving one mile) decreases for larger vehicles; in
contrast, Gramlich42 finds that owners of fuel-inefficient vehicles have the greatest
willingness to pay for improved fuel economy.  Part of the difficulty may be, as these papers
note, that fuel economy may be correlated (either positively or negatively) with other vehicle
attributes, such as size, power, or quality, not all of which may be included in the analyses; as

                                         8-7

-------
Draft Regulatory Impact Analysis

a result, "fuel economy" may in fact represent several characteristics at the same time.
Indeed, Gramlich43 includes both fuel cost (dollars per mile) and miles per gallon in his
analysis, with the argument that miles per gallon measures other undesirable quality
attributes, while fuel cost picks up the consumer's demand for improved fuel economy.

      Espey and Nair44 find, using data from model year 2001, that consumers might be
willing to pay roughly $500 for a 1-mpg increase in city driving, approximately $250 for a 1-
mpg increase in highway driving, or approximately $600 for an increase in combined fuel
economy; they argue that these values approximately correspond to the fuel savings that
consumers might expect over the lifetime of the vehicle.  McManus45 finds, in 2005, that
consumers were willing to pay $578 for a 1-mpg increase in fuel economy. Gramlich46 finds
willingness to pay for an increase from 25 mpg to 30 mpg to range between $4100 (for luxury
cars, when gasoline costs $2/gallon) to $20,600  (for SUVs when gasoline costs $3.50/gallon).

       Some studies47 argue that automakers  could increase profits by increasing fuel
economy because the amount that consumers  are willing to pay for increased fuel economy
outweighs the costs of that improvement. Other studies48 have found that increasing fuel
economy standards imposes welfare losses on consumers and producers, because consumers
should already be buying as much fuel economy as they want. In the course of reaching this
result, though, at least one of these studies49 notes that its baseline model implies that
consumers are willing to buy more fuel economy than producers have provided; they have to
adjust their model to eliminate these "negative-cost" fuel economy improvements.

      The models do not appear to yield very consistent results on the role of fuel economy
in consumer and producer decisions.

    8.1.2.6  Why Consumers May Not Buy, and Producers May Not Provide, Fuel
            Economy that Pays for Itself

      If consumers are willing to pay for fuel-saving technologies, why does the market not
already take advantage of these low-cost technologies? Why aren't consumers demanding
these vehicle improvements, and manufacturers supplying them, when they appear to "pay for
themselves" even in the absence of regulation?

      On the consumer side, this disconnect between net present value estimates of energy-
conserving cost savings and what consumers actually spend on energy conservation is often
referred to as the Energy Paradox,50 since consumers appear to routinely undervalue a wide
range of investments in energy conservation.  Some possible explanations for the paradox
include:

      •  Consumers put little weight on benefits from fuel savings in the future;

      •  Consumers consider other attributes more important than fuel economy at the time
          of vehicle purchase;

      •  Consumers may not be able to find the vehicles they want with improved fuel
          economy;

-------
                                                  Other Economic and Social Impacts

       •  Consumers have difficulty in calculating expected fuel savings;

       •  Consumers may use imprecise rules of thumb when deciding how much fuel
          economy to purchase;

       •  Fuel savings in the future are uncertain;  in contrast, at the time of purchase the
          increased costs of fuel-saving technologies are certain and immediate;

       •  There is likely to be variation among consumers in the benefits they get from
          improved fuel economy, due to different miles driven and driving styles.

       The producer side of this paradox is much less studied. Hypotheses for
underprovision of fuel economy by producers include:

       •  Producers put more effort into attributes that consumers have regularly sought in
          the past, such as size and power, rather than fuel savings with uncertain future
          returns;

       •  In selecting a limited number of vehicle  attributes among which consumers can
          choose, producers may aim to provide choices related to characteristics (such as
          numbers of doors or transmission types) that strongly influence what vehicle a
          consumer will buy, and fuel economy may not make that list;

       •  While consumer preferences for fuel economy may change rapidly as fuel prices
          fluctuate, producers cannot change their design or production decisions as rapidly;
          as a result, vehicle designs may end up not satisfying consumer desires at a
          particular time;

       •  Producers may have misestimated the value that consumers place on fuel
          economy.

       How consumers buy, and producers provide, fuel economy  involves  complex
decisions on both sides of the market. Both sides of the market rely heavily in their
calculations on the uncertain benefits of fuel savings.  In addition, consumers trade off fuel
economy with many other vehicle attributes, and producers do not provide the full range of
attributes possible for consumers.  From this perspective, it may not be a surprise that, at a
given point in time, consumer preferences for fuel economy may not match up with producer
provision of it.

    8.1.2.7   Assessment of the Literature

       Consumer vehicle choice modeling in principle can provide a great deal of useful
information for regulatory analysis. All models estimate changes in fleet mix of new
vehicles; some also provide estimates of total new vehicle sales; and a few incorporate the
used vehicle market, potentially to the decision on when a vehicle is scrapped. Being  able to
model these changes has several advantages.

-------
Draft Regulatory Impact Analysis

       First, consumer vehicle choice modeling has the potential to describe more accurately
the impact of a policy, by identifying market shifts. More accurate description of the market
resulting from a policy can improve other estimates of policy impacts, such as the change in
vehicle emissions or vehicle miles traveled. The predictive ability of models, though, is not
proven. It is likely that, in coming years, new vehicles will be developed, and existing
vehicles will be redesigned, perhaps to have improvements in both fuel economy and safety
factors in combinations that consumers have not previously been offered.  Welch,51 for
instance, argues that auto producers are likely to increase the sizes of vehicles in response to
the footprint-based fuel economy standard.  Models based on the existing vehicle fleet may,
however, not do well in predicting consumers' choices among the new vehicles offered. One
attempt to analyze the effect of the oil shock of 1973 on consumer vehicle choice found that,
after two years, the particular model did not predict well due to changes in the vehicle fleet.52
Thus, consumer vehicle choice models, even if they did produce robust results in analyzing
the short-term effects of policy changes, may miss changes associated with new and
redesigned vehicles.

       The modeling may improve estimates of the compliance costs of a rule. Most current
modeling is based on a fleet mix determined outside the model; neither vehicle manufacturers
nor consumers respond directly to cost increases and other vehicle changes by a change in the
fleet mix. With the use of consumer vehicle choice modeling, both consumers and producers
have greater choices in response to these changes: they can either accept the new costs and
vehicle characteristics, or they can change which vehicles are sold. The fact that consumers
and producers have additional options suggests that compliance costs are likely to be lower
through incorporation of a consumer choice model than through use of a technology-cost
model alone.  On the other hand, the effect may not be large: in the context of "feebates"
(subsidizing fuel-efficient cars with revenue collected by taxing fuel-inefficient vehicles),
Greene et al. found that 95% of the increase in fuel economy was due to addition of
technology rather than changes in vehicles sold.53  Consideration of consumer behavior in
welfare estimates will improve regulatory analysis, but only to the extent that the predicted
changes in consumer purchase patterns reflect actual changes.

       An additional feature of consumer choice models, as noted above, is that they can be
used to calculate consumer surplus impacts. Consumer surplus is a standard measurement of
consumer impacts in benefit-cost analysis. Consumer surplus calculations from these models
estimate how much consumers appreciate the gains in fuel economy relative to the increased
vehicle costs that they face, based on the assumption that consumers,  at the time of vehicle
purchase, have made the best decisions for themselves on the amount of fuel economy in the
vehicles they purchase. These values, though, are based on the relationship between
consumer willingness to pay for fuel economy and the costs of improved fuel economy.
Because the estimates of consumer willingness to pay for fuel economy appear to be highly
inconsistent, consumer surplus measures from any one model are unlikely to be reliable.

       At this point, it is unclear whether two models given the same scenario would produce
similar results in either prediction of changes in the vehicles purchased or in estimates of
consumer surplus effects. The estimates of consumer surplus from consumer vehicle choice
models depend heavily on the value to consumers of improved fuel economy, a value for
which estimates are highly varied. In addition, the predictive ability of consumer vehicle

                                        8-10

-------
                                                  Other Economic and Social Impacts

choice models may be limited as consumers face new vehicle choices that they previously did
not have. If the results across models are not consistent or are highly sensitive to parameters
or other features, then careful thought needs to be given to model selection and development.

       Nonetheless, because there are potential advantages to using consumer vehicle choice
models if these difficulties can be addressed, EPA is continuing to explore options for
including consumer and producer choice in modeling the impacts of fuel economy-related
regulations.  This effort includes further review of existing consumer vehicle choice models
and the estimates of consumers' willingness to pay for increased fuel economy.  In addition,
EPA is developing capacity to examine the factors that may affect the results of consumer
vehicle choice models, and to explore their impact on analysis of regulatory scenarios.  Under
contract with EPA, Resources for the Future (RFF) is developing a model of the  vehicle
market that can be used to evaluate different policy designs and compare regulatory scenarios
on the basis of changes in cost, changes in the prices paid by consumers, changes in consumer
welfare, and changes in industry profits. It should help to shed light on whether  it is more
costly to rely solely on the application of technologies to  vehicles to meet a given fuel
standard than when consumer and producer behavior is taken into account. EPA plans to
evaluate this work within the context of the overall literature on consumer vehicle choices, to
determine its usefulness in informing the analysis for the  final rule. We seek comment on the
usefulness of consumer choice modeling results and the consistency and reliability of results
from these models.
8.1.3 Consumer Payback Period and Lifetime Savings on New Vehicle Purchases

       Another factor of interest is the payback period on the purchase of a new vehicle that
complies with the proposed standards. In other words, how long would it take for the
expected fuel savings to outweigh the increased cost of a new vehicle?  For example, a new
2016 MY vehicle is estimated to cost $1,050 more (on average, and relative to the reference
case vehicle) due to the addition of new GHG reducing technology (see Chapter 4 for details
on this cost estimate).  This new technology will result in lower fuel consumption and,
therefore, savings in fuel expenditures (see Chapter 5 for details on fuel savings). But how
many months or years would pass before the fuel savings exceed the upfront cost of $ 1,050?

       Table 8-3 provides the answer to this question for a vehicle purchaser who pays for the
new vehicle upfront in cash (we discuss later in this section the payback period for consumers
who finance the new vehicle purchase with a loan). The table uses annual miles driven
(vehicle miles traveled, or VMT) and survival rates consistent with the emission and benefits
analyses presented in Chapter 4 of the draft joint TSD. We have included rebound VMT in
the control case but not in the reference case, consistent with other parts of our analysis. We
have also included fuel savings associated with A/C controls (in the control case only), but
have not included expected A/C-related maintenance savings. We discuss the likely
maintenance savings in Chapter 2 of this DRIA. Further, this analysis does not include other
societal impacts such as the value of increased driving, or noise, congestion and accidents
since we really want to focus on those factors consumers consider most while in the
showroom considering a new car purchase.  Car/truck fleet weighting is handled as described

                                         8-11

-------
Draft Regulatory Impact Analysis

in Chapter 1 of the draft joint TSD. As can be seen in the table, it will take under 3 years (2
years and 8 months at a 3% discount rate, 2 years and 10 months at a 7% discount rate) for the
cumulative fuel savings to exceed the upfront increase in vehicle cost. For the average driver,
this payback would occur at around 46,000  to 48,000 miles, depending on the discount rate.
For the driver that drives more than the average, the payback would come sooner.  For the
driver that drives less than the average, the payback would come later.
        Table 8-3 Payback Period on a 2016MY New Vehicle Purchase via Cash (2007 dollars)




Year of
Ownership
1
2
3
4


Increased
Vehicle
Cost"
($)
-$1,128






Fuel
Priceb
($/gal)
$3.27
$3.39
$3.48
$3.56



Reference
VMTC
(miles)
17,481
16,934
16,432
15,777



Control
VMTC
(miles)
17,813
17,256
16,744
16,077


Reference
Fuel
Costs4
($)
$2,544
$2,549
$2,545
$2,495


Control
Fuel
Costs4
($)
$2,101
$2,106
$2,102
$2,061


Annual
Fuel
Savings
($)
$443
$444
$443
$434
Cumulative
Discounted
Fuel
Savings at
3%
($)
$436
$860
$1,272
$1,663
Cumulati
Discount
Fuel
Savings
7%
($)
$428
$829
$1,203
$1,546
a Increased cost of the proposed rule is $1,050; the value here includes nationwide average sales tax of 5.3% and increased insurance
premiums of 1.98%; both of these percentages are discussed in section 8.1.1.
 AEO 2009 reference case fuel price including taxes.
c VMT is calculated as the weighted car/truck VMT with cars estimated to account for 67% of the fleet and trucks 33%; VMT shown
here includes survival fraction and, for the control case, rebound VMT.
 Fuel costs calculated using the reference and control case achieved CO2 levels as presented in Chapter 5 with 8887 grams of CO2 p
gallon of gasoline and include the 20 percent road fuel economy gap, as discussed in Chapter 5; the control case also includes the
effects of A/C controls on CO2 emissions but not the expected A/C-related maintenance savings.

       Most people purchase a new vehicle using credit rather than paying cash up front.  The
typical car loan today is a five year,  60 month loan.  As of August 24, 2009,  the national
average interest rate for a 5 year new car loan was 7.41 percent.  If the increased vehicle cost
is spread out over 5 years at 7.41 percent, the  analysis would look like that shown in Table
8-4.  As can be seen in this table, the fuel savings immediately outweigh the increased
payments on the car loan, amounting to $162  in discounted net savings (3% discount rate)
saved in the first year and similar savings for the next two years before reduced VMT starts to
cause the fuel savings to fall. Results  are similar using a 7% discount rate.  This means that
for every month that the average  owner is making a payment for the financing of the average
new vehicle their monthly fuel savings would be greater than the increase in the loan
payments.  This amounts  to a savings on the order of $9 to $ 14 per month throughout the
duration of the 5 year loan.  Note that in year six when the car loan is paid off, the net
savings equal the fuel savings (as would be the case for the remaining years of ownership).

-------
                                                           Other Economic and Social Impacts
        Table 8-4 Payback Period on a 2016 MY New Vehicle Purchase via Credit (2007 dollars)




Year of
Ownership
1
2
3
4
5
6


Increased
Vehicle
Cost3
($)
$278
$278
$278
$278
$278




Fuel
Priceb
($/gal)
$3.27
$3.39
$3.48
$3.56
$3.62
$3.64



Reference
VMTC
(miles)
17,481
16,934
16,432
15,777
15,109
14,338



Control
VMTC
(miles)
17,813
17,256
16,744
16,077
15,396
14,611


Reference
Fuel
Costs4
($)
$2,544
$2,549
$2,545
$2,495
$2,432
$2,318


Control
Fuel
Costs4
($)
$2,101
$2,106
$2,102
$2,061
$2,009
$1,914


Annual
Fuel
Savings
($)
$443
$444
$443
$434
$423
$403
Annual
Discounted
Net
Savings at
3%
($)
$162
$158
$153
$141
$127
$343
Annua
Discount
Net
Savings
7%
($)
$159
$150
$139
$123
$107
$278
a This uses the same increased cost as Table 8-3 but spreads it out over 5 years assuming a 5 year car loan at 7.41 percent.
 AEO 2009 reference case fuel price including taxes.
c VMT is calculated as the weighted car/truck VMT with cars estimated to account for 67% of the fleet and trucks 33%; VMT shown
here includes survival fraction and, for the control case, rebound VMT.
d Fuel costs calculated using the reference and control case achieved CO2 levels as presented in Chapter 5 with 8887 grams of CO2 p
gallon of gasoline and include the 20 percent road fuel economy gap, as discussed in Chapter 5; the control case also includes the
effects of A/C controls on CO2 emissions but not the expected A/C-related maintenance savings.

  We can also calculate the lifetime fuel savings and net savings for those who purchase the vehicle using
cash and for those who purchase the vehicle with credit.  This calculation applies to the vehicle owner who
    retains the vehicle for its entire life and drives the vehicle each year at the rate equal to the national
projected average. The results are shown in Table  8-5. In either case, the present value of the lifetime net
 savings is greater than $3,200 at a 3% discount rate, or $2,400 at a 7% discount rate.Table 8-5 Lifetime
               Discounted Net Savings on a 2016 MY New Vehicle Purchase (2007 dollars)
Purchase Option
Increased
Discounted Vehicle
Cost
($)
Lifetime
Discounted Fuel
Savings"'0
($)
Lifetime
Discounted Net
Savings
($)
3% discount rate
Cash
Credit3
$1,128
$1,293
$4,558
$4,558
$3,446
$3,265
7% discount rate
Cash
Credit3
$1,128
$1,180
$3,586
$3,586
$2,495
$2,406
            a Assumes a 5 year loan at 7.41 percent.
             VMT is calculated as the weighted car/truck VMT with cars estimated to account for
            67% of the fleet and trucks 33%; VMT shown here includes survival fraction and, for
            the control case, rebound VMT.
            c Fuel  savings here were calculated using AEO 2009 reference case fuel price
            including taxes.

-------
Draft Regulatory Impact Analysis

8.2

8.3 Energy Security Impacts

       This chapter will only describe the energy security analysis that was conducted
beyond that described in Chapter 4.B. 10 of the TSD.  Additional analysis was conducted to
provide inputs to EPA's OMEGA model.  For a detailed discussion of the development of the
energy security estimates, please refer to Chapter 4.B.10 of the TSD.

       After the EPA-sponsored peer review of the Oak Ridge National Laboratory's
(ORNL) Energy Security Analysis was completed in 2008, ORNL, at EPA's request, updated
the analysis using values from the AEO 2009 rather than the 2007 values.  The methodology
used to update this analysis was the same one that was peer-reviewed.54 The results are
shown in Table 8-6. ORNL estimated the energy security premium for 2015, 2020, and 2030.
Since the AEO 2009 forecasts ends in  2030, EPA assumed that the post-2030 energy security
premium did not change through 2040.

          Table 8-6 Energy Security Premium in 2015, 2020, 2030, and 2040 (2007S/Barrel)
YEAR
2015
2020
2030
2040
MONOPSONY
(RANGE)
$11.79
($4.26 -$21. 37)
$12.31
($4.46 - $22.53)
$10.57
($3. 84 -$18.94)
$10.57
($3. 84 -$18.94)
MACROECONOMIC
DISRUPTION/ADJUSTMENT
COSTS (RANGE)
$6.70
($3. 11 -$10.67)
$7.62
($3.77 -$12.46)
$8.12
($3. 90 -$13.04)
$8.12
($3. 90 -$13.04)
TOTAL MID-POINT
(RANGE)
$18.49
($9.80 - $28.08)
$19.94
($10.58 -$30.47)
$18.69
($10.52 - $27.89)
$18.69
($10.52 - $27.89)
       EPA linearly interpolated the values for the years 2016 through 2019, using the 2015
and 2020 values as endpoints. EPA followed the same procedure to estimate the 2021
through 2029 estimates, using the 2020 and 2030 values as endpoints. Post-2030, EPA
assumed that the energy security estimate did not change. The final set of values that was
used by the OMEGA model is shown in Table 8-7.
                                       8-14

-------
                                                  Other Economic and Social Impacts
         Table 8-7 Energy Security Premium Estimates for Years 2015-2040 (2007S/Barrel)
YEAR
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
MONOPSONY
$11.79
$11.89
$12.00
$12.10
$12.21
$12.31
$12.14
$11.96
$11.79
$11.61
$11.44
$11.27
$11.09
$10.92
$10.74
$10.57
$10.57
$10.57
$10.57
$10.57
$10.57
$10.57
$10.57
$10.57
$10.57
$10.57
MACRO/DISRUPT
$6.70
$6.88
$7.07
$7.25
$7.44
$7.62
$7.67
$7.72
$7.77
$7.82
$7.87
$7.92
$7.97
$8.02
$8.07
$8.12
$8.12
$8.12
$8.12
$8.12
$8.12
$8.12
$8.12
$8.12
$8.12
$8.12
TOTAL
$18.49
$18.78
$19.07
$19.36
$19.65
$19.94
$19.82
$19.69
$19.57
$19.44
$19.32
$19.19
$19.07
$18.94
$18.82
$18.69
$18.69
$18.69
$18.69
$18.69
$18.69
$18.69
$18.69
$18.69
$18.69
$18.69
       The total energy security benefits are derived from the estimated reductions in imports
of finished petroleum products and crude oil using only the macroeconomic
disruption/adjustment portion of the energy security premium price.  These values are shown
in Table 8-8.55 The reduced oil estimates were derived from the OMEGA model, as explained
in Chapter 5 of EPA's DRIA. EPA used the same assumption that NHTSA used in its
Corporate Average Fuel Economy and CAFE Reform for MY 2008-2011 Light Trucks
proposal, which assumed each gallon of fuel saved reduces total U.S. imports of crude oil or
refined products by 0.95 gallons56.  Section 5.3 of this RIA contains a discussion regarding
caveats for the fuel savings estimated due to implementation of this rule.  Section III.H.S.b of
the preamble contains a detailed discussion of how the monopsony and macroeconomic
disruption/adjustment components were treated for this analysis. Note that if the monopsony
effects were included in this analysis, they  could be significant.

-------
Draft Regulatory Impact Analysis

 Table 8-8 Total Annual Energy Security Benefits in 2015, 2020, 2030, and 2040 (Billions of 2007 dollars)
YEAR
2015
2020
2030
2040
BENEFITS
$0.59
$2.30
$4.81
$6.23
8.4 Other Externalities

       There are other impacts associated with the proposed GHG emissions standards and
associated reduced fuel consumption. Lower fuel consumption would, presumably, result in
fewer trips to the filling station to refuel and, thus, time saved. The rebound effect, discussed
in detail in Chapter 4 of the draft joint TSD, produces additional benefits to vehicle owners in
the form of consumer surplus from the increase in vehicle-miles driven, but may also increase
the societal costs associated with traffic congestion, motor vehicle crashes, and noise. These
effects are likely to be relatively small in comparison to the value of fuel saved as a result of
the proposed standards, but they are nevertheless important to include. We summarize the
value of these other impacts in section 8.4.4 of this DRIA. Please refer to the draft joint TSD
that accompanies this proposal for more information about these impacts and how EPA and
NHTSA use them in their analyses.

8.4.1 Reduced Refueling Time

       Improving the fuel economy of passenger cars and light-duty trucks may also increase
their driving range before they require refueling.  By reducing the frequency with which
drivers typically refuel their vehicles and extending the upper limit of the range they can
travel before requiring refueling, improving fuel economy provides some additional benefits
to their owners. Alternatively, if manufacturers respond to improved fuel economy by
reducing the size of fuel tanks to maintain a constant driving range, the resulting cost saving
will presumably be reflected in lower vehicle sales prices. If manufacturers respond by doing
so, this presumably reflects their judgment that the value to economic benefits to vehicle
buyers from lower purchase prices exceeds that from extended refueling range.

       No direct estimates of the value of extended vehicle range are readily available, so this
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.57

       Weighted by the nationwide mix of urban (about 2/3) and  rural (about 1/3) driving and
average vehicle occupancy for all driving trips (1.6 persons), the DOT-recommended value of
travel time per vehicle-hour is $24.00 (in 2006 dollars). We assume that the average tank
refill is 55%, that the average fuel tank is 19.3 gallons, and that the average time to find and
use a gas station is five minutes.58'59

-------
                                                   Other Economic and Social Impacts

8.4.2 Value of Additional Driving

       The increase in travel associated with the rebound effect produces additional benefits
to vehicle owners, which reflect the value to drivers and other vehicle occupants of the added
(or more desirable) social and economic opportunities that become accessible with additional
travel. As evidenced by the fact that they elect to make more frequent or longer trips when
the cost of driving declines, the benefits from this added travel exceed drivers' added outlays
for the fuel it consumes (measured at the improved level of fuel economy resulting from
stricter GHG standards ).60  The amount by which the benefits from this increased driving
travel exceed its increased fuel costs measures the net benefits they receive from the
additional travel, usually referred to as increased consumer surplus.

       EPA estimates the economic value of the increased 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

       We discuss the rebound effect in more  detail in Chapter 4 of the draft joint TSD.
Again, the negative effect that rebound driving has on the fuel consumption savings
associated with the proposed GHG standards is included in the fuel economy savings
presented in section 8.5 of this DRIA. Note that in section 8.4.4 below, where we present the
benefit associated with rebound driving, we have used pre-tax fuel prices since those prices
reflect the societal value of the driving.

8.4.3 Noise, Congestion,  and Accidents

       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.  Because drivers do not take these added costs
into account in deciding when and where to travel, they must be accounted  for separately as a
cost of the added driving associated with the rebound effect.

       Increased vehicle use due to the rebound effect may also increase the costs associated
with traffic accidents. Drivers may take account of the potential costs they  (and their
passengers) face from the possibility of being involved in an accident when they decide to
make additional trips. However, they probably do not consider all of the potential costs they
impose on occupants of other vehicles and on pedestrians when accidents occur, so any
increase in these "external" accident costs must be considered as another cost of additional
rebound-effect driving. Like increased delay costs, any increase in these external accident
costs caused by added driving is likely to depend on the traffic conditions under which it takes
place, since accidents are more frequent in heavier traffic (although their severity may be
reduced by the slower speeds at which heavier traffic typically moves).

-------
Draft Regulatory Impact Analysis

       Finally, added vehicle use from the rebound effect may also increase traffic noise.
Noise generated by vehicles causes inconvenience, irritation, and potentially even discomfort
to occupants of other vehicles, to pedestrians and other bystanders, and to residents or
occupants of surrounding property. Because these effects are unlikely to be taken into
account by the drivers whose vehicles contribute to traffic noise, they represent additional
externalities associated with motor vehicle use. Although there is considerable uncertainty in
measuring their value, any increase in the economic costs of traffic noise resulting from added
vehicle use must be included together with other increased external costs from the rebound
effect.

       EPA relies 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.61  NHTSA
employed these estimates previously in its analysis accompanying the MY 2011 final rule,
and continues to find them appropriate for this analysis after reviewing the procedures used
by FHWA to develop them and considering other available estimates of these values. They
are intended to measure the increases in costs from added congestion, property damages and
injuries in traffic accidents, and noise levels caused by automobiles and light trucks that are
borne by persons other than their drivers (or "marginal" external costs).

       Updated to 2007 dollars, FHWA's "Middle" estimates for marginal congestion,
accident, and noise costs caused by automobile use amount to 5.2 cents, 2.3 cents, and 0.1
cents per vehicle-mile (for a total of 7.6 cents per mile), while those for pickup trucks and
vans are 4.7 cents, 2.5 cents, and 0.1 cents per vehicle-mile (for a total of 7.3 cents per
mile).62'63  These costs are multiplied by the annual increases in automobile and light truck
use from the rebound effect to yield the estimated increases in congestion,  accident, and noise
externality costs during each future year.

       EPA uses a single value for both cars and trucks, as shown in Table 8-9.

                        Table 8-9 S/mile Inputs used for External Costs
EXTERNAL COSTS
Congestion
Accidents
Noise
$/VMT
$ 0.052
$ 0.023
$ 0.001
8.4.4 Summary of Other Externalities

       Table 8-10 summarizes the other economic impacts discussed in sections 8.4.1
through 8.4.3.

-------
                                                   Other Economic and Social Impacts
Table 8-10. Estimated Economic Externalities Associated with the Proposed Light-Duty Vehicle GHG
                            Program (Millions of 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
REDUCED
REFUELING
$100
$200
$500
$700
$1,100
$1,500
$1,800
$2,200
$2,500
$2,800
$3,100
$3,400
$3,700
$3,900
$4,200
$4,400
$4,600
$4,800
$4,900
$5,100
$5,200
$5,400
$5,500
$5,700
$5,800
$6,000
$6,100
$6,200
$6,400
$6,500
$6,700
$6,800
$7,000
$7,200
$7,300
$7,500
$7,700
$7,800
$8,000
$89,600
$41,000
VALUE OF
INCREASED
DRIVING
$100
$400
$700
$1,200
$2,000
$2,700
$3,400
$4,200
$4,900
$5,500
$6,100
$6,700
$7,200
$7,700
$8,200
$8,600
$9,100
$9,800
$10,000
$10,400
$10,700
$11,100
$11,400
$11,800
$12,100
$12,500
$12,900
$13,200
$13,600
$14,000
$14,400
$14,800
$15,200
$15,700
$16,100
$16,600
$17,000
$17,500
$18,000
$184,700
$82,700
ACCIDENTS,
NOISE,
CONGESTION
-$100
-$200
-$400
-$700
-$1,100
-$1,400
-$1,800
-$2,100
-$2,400
-$2,700
-$3,000
-$3,300
-$3,600
-$3,900
-$4,100
-$4,300
-$4,500
-$4,700
-$4,900
-$5,000
-$5,200
-$5,300
-$5,500
-$5,600
-$5,800
-$5,900
-$6,000
-$6,200
-$6,300
-$6,500
-$6,600
-$6,800
-$6,900
-$7,100
-$7,200
-$7,400
-$7,600
-$7,800
-$7,900
-$88,200
-$40,200
ANNUAL
QUANTIFIED
BENEFITS
$200
$400
$800
$1,300
$2,000
$2,800
$3,500
$4,200
$5,000
$5,600
$6,200
$6,800
$7,300
$7,800
$8,300
$8,700
$9,200
$9,800
$10,000
$10,400
$10,800
$11,100
$11,500
$11,800
$12,200
$12,600
$12,900
$13,300
$13,700
$14,100
$14,500
$14,900
$15,300
$15,700
$16,200
$16,700
$17,100
$17,600
$18,100
$186,100
$83,500

-------
Draft Regulatory Impact Analysis

8.5 Summary of Costs and Benefits

       In this section we present a summary of costs, benefits, and net benefits of the
proposal. We present fuel consumption impacts as negative costs of the vehicle program.

       Table 8-11 shows the estimated annual societal costs of the vehicle program for the
indicated calendar years. The table also shows the net present values of those costs for the
calendar years 2012-2050 using both a 3 percent and a seven percent discount rate.  In this
table, fuel savings are calculated using pre-tax fuel prices and are presented as negative costs
associated with the vehicle program (rather than positive savings).

 Table 8-11 Estimated Societal Costs of the Light-Duty Vehicle GHG  Program (Millions of 2007 dollars)
COSTS
Vehicle Compliance Costs
Fuel Savings a
Quantified Annual Costs
2020
$18,000
-$43,100
-$25,100
2030
$17,900
-$90,400
-$72,500
2040
$19,300
-$125,000
-$105,700
2050
$20,900
-$167,000
-$146,100
NPV, 3%
$390,000
-$1,677,600
-$1,287,600
NPV, 7%
$216,600
-$746,100
-$529,500
a Calculated using pre-tax fuel prices.

       Table 8-12 presents estimated annual societal benefits for the indicated calendar years.
The table also shows the net present values of those benefits for the calendar years 2012-2050
using both a 3 percent and a 7 percent discount rate. The table shows the benefits of reduced
GHG emissions—and consequently the annual quantified benefits (i.e., total benefits)—for
each of five interim SCC values considered by EPA.

       The interim SCC values are derived using several discount rates and include:

          •   $5 (based on a 5% discount rate);

          •   $ 10  (5% using Newell-Pizer adjustment);,

          •   $20  (average SCC value from the average SCC estimates based on 5% and
              3%);

          •   $34

          •   $56  (3% using Newell-Pizer adjustment).

These interim SCC values are in 2007 dollars, and are based on a CCh emissions change of 1
metric ton in 2007.  Section III.H.2.a of the Preamble provides a complete discussion about
SCC and the interim set of values.

       Section III.H.2.a of the preamble to this rule also notes that there is a very high
probability (very likely according to the IPCC) that the benefit estimates from GHG
reductions are underestimates. One of the primary reasons is that models used to calculate
SCC values do not include information about impacts that have not been quantified.

-------
                                                        Other Economic and Social Impacts
        In addition, the total GHG reduction benefits presented below likely underestimate the
value of GHG reductions because they were calculated using the marginal values for CCh
emissions.  The impacts of non-CCh emissions vary from those of CCh emissions because of
differences in atmospheric lifetimes and radiative forcing.  As a result, the marginal benefit
values of non-CO2 GHG reductions and their growth rates over time will not be the same as
the marginal benefits measured on a CCh-equivalent scale.  Marginal benefit estimates per
metric ton of non- CCh GHGs are currently unavailable, but work is on-going to monetize
benefits related to the mitigation of other non-CCh GHGs.

    Table 8-12 Use Estimated Societal Benefits Associated with the Proposed Light-Duty Vehicle GHG
                                Program (Millions of 2007 dollars)
BENEFITS
2020
2030
2040
2050
NPV, 3%
NPV, 7%
Reduced GHG Emissions at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
PM2.5 Related Benefits3""'0
Energy Security Impacts
(price shock)
Reduced Refueling
Value of Increased Driving"
Accidents, Noise,
Congestion
$1,200
$2,500
$4,700
$8,200
$14,000
$1,400
$2,300
$2,500
$4,900
-$2,400
$3,300
$6,600
$12,000
$22,000
$36,000
$3,000
$4,800
$4,900
$10,000
-$4,900
$5,700
$11,000
$22,000
$38,000
$63,000
$4,600
$6,200
$6,400
$13,600
-$6,300
$9,500
$19,000
$36,000
$63,000
$100,000
$6,700
$7,800
$8,000
$18,000
-$7,900
$69,200
$138,400
$263,000
$456,900
$761,400
$59,800
$85,800
$89,600
$184,700
-$88,200
$28,600
$57,100
$108,500
$188,500
$314,200
$26,300
$38,800
$41,000
$82,700
-$40,200
Quantified Annual Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$9,900
$11,200
$13,400
$16,900
$22,700
$21,100
$24,400
$29,800
$39,800
$53,800
$30,200
$35,500
$46,500
$62,500
$87,500
$42,100
$51,600
$68,600
$95,600
$132,600
$400,900
$470,100
$594,700
$788,600
$1,093,100
$177,200
$205,700
$257,100
$337,100
$462,800
aNote that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts. Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only human
health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental benefits
would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality modeling.
However, we were unable to conduct a full-scale air quality modeling analysis in time for the proposal.  We
intend to more fully capture the co-pollutant benefits for the analysis of the final standards.
 The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table are based on an
estimate of premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates
were based on the Six Cities study (Laden et al., 2006), the values would be approximately 145% (nearly two-
and-a-half times) larger
c The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table assume a 3% discount
rate in the valuation of premature  mortality to account for a twenty-year segmented cessation lag.  If a 7%
discount rate had been used, the values would be approximately 9% lower.
 Calculated using pre-tax fuel prices.
        Table 8-13 presents estimated annual net benefits for the indicated calendar years.
The table also shows the net present values of those net benefits for the calendar years 2012-

-------
Draft Regulatory Impact Analysis
2050 using both a 3 percent and a 7 percent discount rate.  The table includes the benefits of
reduced GHG emissions—and consequently the annual net benefits—for each of five interim
SCC values considered by EPA.  As noted above, there is a very high probability (very likely
according to the IPCC) that the benefit estimates from GHG reductions are underestimates
because, in part, models used to calculate SCC values do not include information about
impacts that have not been quantified.

 Table 8-13. Quantified Net Benefits Associated with the Proposed Light-Duty Vehicle GHG Program3'b
                                  (Millions of 2007 dollars)

Quantified Annual
Costs
2020
-$25,100
2030
-$72,500
2040
-$105,700
2050
-$146,100
NPV, 3%
-$1,287,600
NPV, 7%
-$529,500
Quantified Annual Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$9,900
$11,200
$13,400
$16,900
$22,700
$21,100
$24,400
$29,800
$39,800
$53,800
$30,200
$35,500
$46,500
$62,500
$87,500
$42,100
$51,600
$68,600
$95,600
$132,600
$400,900
$470,100
$594,700
$788,600
$1,093,100
$177,200
$205,700
$257,100
$337,100
$462,800
Quantified Net Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$35,000
$36,300
$38,500
$42,000
$47,800
$93,600
$96,900
$102,300
$112,300
$126,300
$135,900
$141,200
$152,200
$168,200
$193,200
$188,200
$197,700
$214,700
$241,700
$278,700
$1,688,500
$1,757,700
$1,882,300
$2,076,200
$2,380,700
$706,700
$735,200
$786,600
$866,600
$992,300
"Note that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts.  Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only human
health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental benefits
would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality modeling.
However, we were unable to conduct a full-scale air quality modeling analysis in time for the proposal.  We
intend to more fully capture the co-pollutant benefits for the analysis of the final standards.
bFuel impacts were calculated using pre-tax fuel prices.
       EPA also conducted a separate analysis of the total benefits over the model year
lifetimes of the 2012 through 2016 model year vehicles. In contrast to the calendar year
analysis, the model year lifetime analysis shows the lifetime impacts of the program on each
of these MY fleets over the course of its lifetime.  Full details of the inputs to this analysis can
be found in DRIA chapter 5. The societal benefits of the full life of each of the five model
years from 2012 through 2016 are shown in Table 8-14 and Table 8-15 at both a 3 percent and
a 7 percent discount rate, respectively.  The net benefits are shown in Table 8-16 and Table
8-17 for both a 3  percent and a 7 percent discount rate, respectively.  Note that the quantified
annual benefits shown in Table 8-14 and Table 8-15include fuel savings as a positive benefit.
As such, the quantified annual costs as  shown in Table 8-16  and Table 8-17 do not include
fuel savings since those are included as benefits. Also note that Table 8-14 through Table
8-17 include the benefits of reduced CCh emissions—and consequently the total benefits—for
each of five interim SCC values considered by EPA. As noted above, there is a very high

-------
                                                               Other Economic and Social Impacts
probability (very likely according to the IPCC) that the benefit estimates from GHG
reductions are underestimates because, in part, models used to calculate SCC values do not
include information about impacts that have not been quantified.

 Table 8-14 Estimated Societal Benefits Associated with the Proposed Light-Duty Vehicle GHG Program,
                    Model Year Analysis (Millions of 2007  dollars; 3% Discount Rate)
MONETIZED VALUES
Cost of Noise, Accident,
Congestion ($)
Pretax Fuel Savings ($)
Energy Security ($) (price
shock)
Change in no. of Refueling
(#)
Change in Refueling Time
(hours)
Value of Reduced Refueling
time ($)
Value of Additional Driving
($)
Value of PM2.5 related Health
Impacts ($)a'b'c
2012MY
-$900
$15,600
$400
500
0
$900
$2,000
$600
2013MY
-$1,400
$24,400
$600
700
100
$1,400
$3,000
$900
2014MY
-$1,900
$34,800
$900
1,000
100
$1,900
$4,100
$1,200
2015MY
-$2,800
$49,800
$1,200
1,300
100
$2,700
$5,700
$1,700
2016MY
-$3,900
$68,500
$1,600
1,800
200
$3,700
$7,900
$2,200
SUM
-$11,000
$193,300
$4,700
5,300
400
$10,500
$22,700
$6,600
Social Cost of Carbon (SCC) at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$500
$1,000
$1,800
$3,200
$5,300
$700
$1,500
$2,800
$4,800
$8,100
$1,000
$2,000
$3,900
$6,700
$11,000
$1,400
$2,900
$5,400
$9,400
$16,000
$1,900
$3,800
$7,200
$13,000
$21,000
$5,600
$11,000
$21,000
$37,000
$61,000
Total Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$19,100
$19,600
$20,400
$21,800
$23,900
$29,600
$30,400
$31,700
$33,700
$37,000
$42,000
$43,000
$44,900
$47,700
$52,000
$59,700
$61,200
$63,700
$67,700
$74,300
$81,900
$83,800
$87,200
$93,000
$101,000
$232,400
$237,800
$247,800
$263,800
$287,800
3 Note that the co-pollutant impacts associated with the standards presented here do not include the full complement of
endpoints that, if quantified and monetized, would change the total monetized estimate of rule-related impacts.  Instead, the
co-pollutant benefits are based on benefit-per-ton values that reflect only human health impacts associated with reductions in
PM2.5 exposure. Ideally, human health and environmental benefits would be based on changes in ambient PM2.5 and ozone
as determined by full-scale air quality modeling.  However, we were unable to conduct a full-scale air quality modeling
analysis in time for the proposal. We intend to more fully capture the co-pollutant benefits for the analysis of the final
standards.
b The PM2.5-related benefits (derived from benefit-per-ton values) presented in mis table are based on an estimate of
premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six
Cities study (Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
c The PM2.5-related benefits (derived from benefit-per-ton values) presented in mis table assume a 3% discount rate in the
valuation of premature mortality to account for a twenty-year segmented cessation lag. If a 7% discount rate had been used,
the values would be approximately 9% lower.

-------
Draft Regulatory Impact Analysis
 Table 8-15. Estimated Societal Benefits Associated with the Proposed Light-Duty Vehicle GHG Program,
                     Model Year Analysis (Millions of 2007 dollars; 7% Discount Rate)
MONETIZED VALUES
Cost of Noise, Accident,
Congestion ($)
Pretax Fuel Savings (S)
Energy Security (S) (price
shock)
Change in no. of Refueling (#)
Change in Refueling Time
(hours)
Value of Reduced Refueling
time ($)
Value of Additional Driving
($)
Value of PM2.5 related Health
Impacts ($)a'b'c
2012MY
-$700
$12,100
$300
400
0
$700
$1,500
$500
2013MY
-$1,100
$19,000
$500
500
0
$1,100
$2,400
$700
2014MY
-$1,500
$27,200
$700
800
100
$1,500
$3,200
$1,000
2015MY
-$2,200
$39,000
$900
1,100
100
$2,100
$4,500
$1,300
2016MY
-$3,100
$53,700
$1,300
1,500
100
$2,900
$6,300
$1,800
SUM
-$8,700
$150,900
$3,700
4,200
300
$8,300
$18,000
$5,300
Social Cost of Carbon (SCC) at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$400
$700
$1,400
$2,400
$4,000
$500
$1,100
$2,100
$3,600
$6,000
$700
$1,500
$2,800
$4,800
$8,000
$1,000
$2,000
$3,700
$6,500
$11,000
$1,300
$2,500
$4,800
$8,300
$14,000
$3,900
$7,700
$15,000
$26,000
$43,000
Total Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$14,800
$15,100
$15,800
$16,800
$18,400
$23,100
$23,700
$24,700
$26,200
$28,600
$32,800
$33,600
$34,900
$36,900
$40,100
$46,600
$47,600
$49,300
$52,100
$56,600
$64,200
$65,400
$67,700
$71,200
$76,900
$181,400
$185,200
$192,500
$203,500
$220,500
aNote that the co-pollutant impacts associated with the standards presented here do not include the full complement of
endpoints that, if quantified and monetized, would change the total monetized estimate of rule-related impacts.  Instead, the
co-pollutant benefits are based on benefit-per-ton values that reflect only human health impacts associated with reductions in
PM2.5 exposure. Ideally, human health and environmental benefits would be based on changes in ambient PM2.5 and ozone
as determined by full-scale air quality modeling. However, we were unable to conduct a full-scale air quality modeling
analysis in time for the proposal. We intend to more fully capture the co-pollutant benefits for the analysis of the final
standards.
b The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table are based on an estimate of
premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six
Cities study (Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
c The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table assume a 3% discount rate in the
valuation of premature mortality to account for a twenty-year segmented cessation lag. If a 7% discount rate had been used,
the values would be approximately 9% lower.

-------
                                                           Other Economic and Social Impacts
  Table 8-16. Quantified Net Benefits Associated with the Proposed Light-Duty Vehicle GHG Program,
                   Model Year Analysis3 (Millions of 2007 dollars; 3% Discount Rate)

Quantified Annual Costs
(excluding fuel savings)
2012MY
$5,400
2013MY
$8,400
2014MY
$10,900
2015MY
$13,900
2016MY
$17,500
SUM
$56,100
Quantified Annual Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$19,100
$19,600
$20,400
$21,800
$23,900
$29,600
$30,400
$31,700
$33,700
$37,000
$42,000
$43,000
$44,900
$47,700
$52,000
$59,700
$61,200
$63,700
$67,700
$74,300
$81,900
$83,800
$87,200
$93,000
$101,000
$232,400
$237,800
$247,800
$263,800
$287,800
Quantified Net Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$13,700
$14,200
$15,000
$16,400
$18,500
$21,200
$22,000
$23,300
$25,300
$28,600
$31,100
$32,100
$34,000
$36,800
$41,100
$45,800
$47,300
$49,800
$53,800
$60,400
$64,400
$66,300
$69,700
$75,500
$83,500
$176,300
$181,700
$191,700
$207,700
$231,700
"Note that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts. Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only human
health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental benefits
would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality modeling.
However, we were unable to conduct a full-scale air quality modeling analysis in time for the proposal.  We
intend to more fully capture the co-pollutant benefits for the analysis of the final standards.

  Table 8-17. Quantified Net Benefits Associated with the Proposed Light-Duty Vehicle GHG Program,
                  Model Year Analysis3 (Millions of 2007 dollars; 7% Discount Rate)

Quantified Annual Costs
(excluding fuel savings)
2012MY
$5,400
2013MY
$8,400
2014MY
$10,900
2015MY
$13,900
2016MY
$17,500
SUM
$56,100
Quantified Annual Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$14,800
$15,100
$15,800
$16,800
$18,400
$23,100
$23,700
$24,700
$26,200
$28,600
$32,800
$33,600
$34,900
$36,900
$40,100
$46,600
$47,600
$49,300
$52,100
$56,600
$64,200
$65,400
$67,700
$71,200
$76,900
$181,400
$185,200
$192,500
$203,500
$220,500
Quantified Net Benefits at each assumed SCC value
SCC 5%
SCC5%Newell-Pizer
SCC from 3% and 5%
SCC 3%
SCC3%Newell-Pizer
$9,400
$9,700
$10,400
$11,400
$13,000
$14,700
$15,300
$16,300
$17,800
$20,200
$21,900
$22,700
$24,000
$26,000
$29,200
$32,700
$33,700
$35,400
$38,200
$42,700
$46,700
$47,900
$50,200
$53,700
$59,400
$125,300
$129,100
$136,400
$147,400
$164,400
aNote that the co-pollutant impacts associated with the standards presented here do not include the full
complement of endpoints that, if quantified and monetized, would change the total monetized estimate of rule-
related impacts. Instead, the co-pollutant benefits are based on benefit-per-ton values that reflect only human
health impacts associated with reductions in PM2.5 exposure. Ideally, human health and environmental benefits
would be based on changes in ambient PM2.5 and ozone as determined by full-scale air quality modeling.
However, we were unable to conduct a full-scale air quality modeling analysis in time for the proposal.  We
intend to more fully capture the co-pollutant benefits for the analysis of the final standards.

-------
Draft Regulatory Impact Analysis


        References

        All references can be found in the EPA DOCKET: EPA-HQ-OAR-2009-0472.
1 See, for instance, Gron, Ann, and Deborah Swenson, 2000. "Cost Pass-Through in the U.S. Automobile
Market," Review of Economics and Statistics 82: 316-324.

2 Insurance Information Institute, 2008, "Average Expenditures for Auto Insurance By State, 2005-2006,"
http://www.iii.org/media/facts/statsbyissue/auto/, accessed April 23, 2009.

3 U.S. Department of Energy, 2008, "Average Price of aNew Car, 1970-2006,"
http://wwwl.eere.energy.gov/vehiclesandfuels/facts/2008_fotw520.html, accessed April 23, 2009.

4 Solheim, Mark, 2006"State Car Tax Rankings,"
http://www.kiplinger.com/features/archives/2006/04/cartax.html, accessed April 23, 2009.

5 U.S. Census Bureau, "Population, Population change and estimated components of population change: April 1,
2000 to July 1, 2008" (NST-EST2008-alldata), http://www.census.gov/popest/states/states.html, accessed April
23, 2009.

6 http://auto-loan.interest.com. 8/24/09 at 9:20AM.

7 Consumer Reports, August 2008,"What That Car Really Costs to Own,"
http://www.consumerreports.org/cro/cars/pricing/what-that-car-really-costs-to-own-4-08/overview/what-that-
car-really-costs-to-own-ov.htm , accessed April 23, 2009.

8 Kleit A.N., 1990, "The Effect of Annual Changes in Automobile Fuel Economy Standards,"  Journal of
Regulatory Economics 2: 151-172;  McCarthy, Patrick S., 1996. "Market Price and Income Elasticities of New
Vehicle Demands." Review of Economics and Statistics 78: 543-547; Goldberg, Pinelopi  K., 1998. "The Effects
of the Corporate Average Fuel Efficiency Standards in the US," Journal of Industrial Economics 46(1): 1-33;
Greene, David L., "Feebates, Footprints and Highway Safety," Transportation Research Part D 14 (2009): 375-
384.

9 Goldberg, Pinelopi Koujianou, "Product Differentiation and Oligopoly in International Markets: The Case of
the U.S. Automobile Industry," Econometrica 63(4) (July 1995):  891-951; Goldberg, Pinelopi Koujianou, "The
Effects of the Corporate Average Fuel Efficiency Standards in the US," Journal of Industrial Economics 46(1)
(March 1998): 1-33.

10 Greene, David L., K.G. Duleep, Doug Elliott, and Sanjana Ahmad, "A Fuel Economy Regulatory Analysis
Model (FERAM) For the Energy Information Administration," prepared by the Oak Ridge National Laboratory
for the U.S. Department of Energy under contract No. DE-AC0500OR22725, 2005; Greene, David L., Philip D.
Patterson, Margaret Singh, and Jia Li, "Feebates, Rebates, and Gas-Guzzler Taxes: A Study of Incentives for
Increased Fuel Economy," Energy Policy 33  (2005): 757-775.

11 McManus, Walter M., "Can Proactive Fuel Economy Strategies Help Automakers Mitigate Fuel-Price Risks?"
University of Michigan Transportation Research Institute, September 14, 2006.

12 Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market Equilibrium," Econometrica
63(4) (July 1995): 841-940; Berry, Steven, James Levinsohn, and Ariel Pakes, "Differentiated Products Demand
Systems from a Combination of Micro and Macro Data:  The New Car Market," Journal of Political Economy
112(1) (2004): 68-105.

-------
                                                          Other Economic and Social Impacts
13 Bento, Antonio M., Lawrence H. Goulder, Emeric Henry, Mark R. Jacobsen, and Roger H. von Haefen,
"Distributional and Efficiency Impacts of Gasoline Taxes: An Econometrically Based Multi-Market Study,"
American Economic Review 95(2) (May 2005): 282-287.

14 Train, Kenneth E., and Clifford Winston, "Vehicle Choice Behavior and the Declining Market Share of U.S.
Automakers," International Economic Review48(4) (November 2007):  1469-1496.

  Kleit, Andrew N., "Impacts of Long-Range Increases in the Fuel Economy (CAFE) Standard," Economic
Inquiry 42(2) (April 2004):  279-294.

16 Austin, David, and Terry Dinan, "Clearing the Air: The Costs and Consequences of Higher CAFE Standards
and Increased Gasoline Taxes," Journal of Environmental Economics and Management 50 (2005): 562-582.

17 E.g., Bento, Antonio M., Lawrence H. Goulder, Emeric Henry, Mark R. Jacobsen, and Roger H. von Haefen,
"Distributional and Efficiency Impacts of Gasoline Taxes: An Econometrically Based Multi-Market Study,"
American Economic Review 95(2) (May 2005): 282-287; Train, Kenneth E., and Clifford Winston, "Vehicle
Choice Behavior and the Declining Market Share of U.S. Automakers," International Economic Review 48(4)
(November 2007): 1469-1496.

18 E.g., Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market Equilibrium,"
Econometrica 63(4) (July 1995):  841-940.

19 Greene, David L., K.G. Duleep, Doug Elliott, and Sanjana Ahmad, "A Fuel Economy Regulatory Analysis
Model (FERAM) For the Energy Information Administration," prepared by the Oak Ridge National Laboratory
for the U.S. Department of Energy under contract No. DE-AC0500OR22725, 2005.

20NERA Economic Consulting, "Evaluation ofNHTSA's Benefit-Cost Analysis of 2011-2015 CAFE
Standards," 2008, available at
http://www.heartland.org/policybot/results/23495/Evaluation_of_NHTSAs_BenefitCost_Analysis_Of_20112015
_CAFE_Standards.html.

21 E.g., Train, Kenneth E., and Clifford Winston, "Vehicle Choice Behavior and the Declining Market  Share of
U.S. Automakers." International Economic Review 48(4) (November 2007): 1469-1496.

  E.g., Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market Equilibrium,"
Econometrica 63(4) (July 1995):  841-940.

23 E.g., Bento, Antonio M., Lawrence H. Goulder, Emeric Henry, Mark R. Jacobsen, and Roger H. von Haefen,
"Distributional and Efficiency Impacts of Gasoline Taxes: An Econometrically Based Multi-Market Study,"
American Economic Review 95(2) (May 2005): 282-287.

24 Bento, Antonio M., Lawrence H. Goulder, Emeric Henry, Mark R. Jacobsen, and Roger H. von Haefen,
"Distributional and Efficiency Impacts of Gasoline Taxes: An Econometrically Based Multi-Market Study,"
American Economic Review 95(2) (May 2005): 282-287; Feng, Yi, Don Fullerton, and Li Gan, "Vehicle
Choices, Miles Driven and Pollution Policies," National Bureau of Economic Analysis Working Paper 11553,
available at http://econweb.tamu.edu/gan/wll553.pdf, accessed 5/12/09.

25 NERA Economic Consulting, "Appendix B: New Vehicle Market Model," "Impacts of the California
Greenhouse Gas Emission Standards on Motor Vehicle Sales," comments submitted to the U.S. Environmental
Protection Agency at Regulations.gov, document number EPA-HQ-OAR-2006-0173-9053.1.

26 Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market Equilibrium," Econometrica
63(4)  (July 1995): 841-940.
                                               8-27

-------
Draft Regulatory Impact Analysis
  Goldberg, Pinelopi Koujianou, "Product Differentiation and Oligopoly in International Markets:  The Case of
the U.S. Automobile Industry." Econometrica 63(4) (July 1995): 891-951.

28 Brownstone, David, and Kenneth Train, "Forecasting New Product Penetration with Flexible Substitution
Patterns," Journal of Econometrics 89 (1999):  109-129; Brownstone, David, David S. Bunch, and Kenneth
Train, "Joint Mixed Logit Models of Stated and Revealed Preferences for Alternative-Fuel Vehicles,"
Transportation Research Part B 34 (2000):  315-338; Greene David L., "TAFV Alternative Fuels and Vehicles
Choice Model Documentation," prepared by the Oak Ridge National Laboratory for the U.S. Department of
Energy, July 2001; Greene, David L., K. G. Duleep, and Walter McManus, "Future Potential of Hybrid and
Diesel Powertrains in the U.S.  Light-Duty Vehicle Market," prepared by the Oak Ridge National Laboratory for
the U.S. Department of Energy, August 2004.

29 Petrin, Amil, "Quantifying the Benefits of New Products:  The Case of the Minivan," Journal of Political
Economy 110 (2002): 705-729; Berry,  Steven, James Levinsohn, and Ariel Pakes, "Differentiated Products
Demand Systems from a Combination of Micro and Macro Data: The New Car Market," Journal of Political
Economy 112 (2004): 68-105.

30 Train, Kenneth E., and Clifford Winston, "Vehicle Choice Behavior and the Declining Market Share of U.S.
Automakers," International Economic Review 48 (November 2007):  1469-1496.

31 Bento, Antonio M., Lawrence H. Goulder, Emeric Henry, Mark R. Jacobsen, and Roger H. von Haefen,
"Distributional and Efficiency  Impacts of Gasoline Taxes: An Econometrically Based Multi-Market Study,"
American Economic Review 95(2) (May 2005): 282-287; Feng, Yi, Don Fullerton, and Li Gan, "Vehicle
Choices, Miles Driven and Pollution Policies," National Bureau of Economic Analysis Working Paper 11553,
available at http://econweb.tamu.edu/gan/wll553.pdf, accessed 5/12/09.

32 Greene, David L., Philip D. Patterson, Margaret Singh, and Jia Li, "Feebates, Rebates, and Gas-Guzzler
Taxes:  A Study of Incentives for Increased Fuel Economy," Energy Policy 33 (2005): 757-775; Feng, Yi, Don
Fullerton, and Li Gan, "Vehicle Choices, Miles Driven and Pollution Policies," National Bureau of Economic
Analysis Working Paper 11553, available at http://econweb.tamu.edu/gan/wl 1553.pdf, accessed 5/12/09;
Greene, David L., "Feebates, Footprints and Highway Safety," Transportation Research Part D (2009) (in press).

33 E.g., Austin, David, and Terry Dinan, "Clearing the Air: The Costs and Consequences of Higher CAFE
Standards and Increased Gasoline Taxes," Journal of Environmental Economics and Management 50 (2005):
562-582.

34 Turrentine, Thomas S., and Kenneth S. Kurani, "Car Buyers and Fuel Economy?" Energy Policy 35 (2007):
1213-1223.

35 E.g., Espey, Molly, and Santosh Nair, 2005.  "Automobile Fuel Economy:  What Is It Worth?" Contemporary
Economic Policy 23: 317-323.
36
  Larrick, Richard P., and Jack B. Soil, 2008. "The MPG Illusion," Science 320(5883):  1593-1594.
37 E.g., Goldberg, Pinelopi Koujianou, "Product Differentiation and Oligopoly in International Markets:  The
Case oftheU.S. Automobile Industry," Econometrica 63(4) (July 1995): 891-951.

38 E.g., Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market Equilibrium,"
Econometrica 63(4) (July 1995): 841-940.

39 Busse, Meghan, Christopher R. Knittel, and Florian Zettelmeyer, 2009. "Pain at the Pump:  How Gasoline
Prices Affect Automobile Purchasing in New and Used Markets," working paper,
http://www.econ.ucdavis.edu/faculty/knittel/papers/gaspaper_latest.pdf (accessed 6/16/09).
                                                8-28

-------
                                                           Other Economic and Social Impacts
40 Greene, David L., and Jin-Tan Liu, 1988. "Automotive Fuel Economy Improvements and Consumers'
Surplus,"  Transportation Research Part A 22A (3): 203-218.

41 Goldberg, Pinelopi Koujianou, "Product Differentiation and Oligopoly in International Markets: The Case of
the U.S. Automobile Industry," Econometrica 63(4) (July 1995):  891-951; Berry, Steven, James Levinsohn, and
Ariel Pakes, "Automobile Prices in Market Equilibrium," Econometrica 63(4) (July 1995):  841-940.

  Gramlich, Jacob, "Gas Prices and Endogenous Product Selection in the U.S. Automobile Industry,"
http://www.econ.yale.edu/seminars/apmicro/am08/gramlich-081216.pdf, accessed 5/11/09.

43 Gramlich, Jacob, "Gas Prices and Endogenous Product Selection in the U.S. Automobile Industry,"
http://www.econ.yale.edu/seminars/apmicro/am08/gramlich-081216.pdf, accessed 5/11/09.

44 Espey, Molly, and Santosh Nair, 2005. "Automobile Fuel Economy:  What Is It Worth?" Contemporary
Economic Policy 23: 317-323.

45 McManus, Walter, "Can Proactive Fuel Economy Strategies Help Automakers Mitigate Fuel-Price Risks?"
Ann Arbor, MI: University of Michigan Transportation Research Institute; Munich Personal RePEc Archive
Paper No. 3460, September 14,2006; http://mpra.ub.uni-muenchen.de/3460/l/MPRA_paper_3460.pdf, accessed
6/16/09.

46 Gramlich, Jacob, "Gas Prices and Endogenous Product Selection in the U.S. Automobile Industry,"
http://www.econ.yale.edu/seminars/apmicro/am08/gramlich-081216.pdf, accessed 5/11/09.

47 Gramlich, Jacob, "Gas Prices and Endogenous Product Selection in the U.S. Automobile Industry,"
http://www.econ.yale.edu/seminars/apmicro/am08/gramlich-081216.pdf, accessed 5/11/09; McManus, Walter,
2007. "The Impact of Attribute-Based Corporate Average Fuel Economy (CAFE) Standards:  Preliminary
Results." Ann Arbor, MI: University of Michigan Transportation Research Institute, Report No. UMTRI-2007-
31.

48 For instance, Kleit, Andrew N., "Impacts of Long-Range Increases in the Fuel Economy (CAFE) Standard,"
Economic Inquiry 42(2) (April 2004):  279-294; Austin, David, and Terry Dinan, "Clearing the Air:  The Costs
and Consequences of Higher CAFE Standards and Increased Gasoline Taxes," Journal of En vironmental
Economics and Management 50 (2005): 562-582; Klier, Thomas, and Joshua Linn, 2008.  "New Vehicle
Characteristics and the Cost of the Corporate Average Fuel Economy Standard," Federal Reserve Bank of
Chicago WP 2008-13, at http://www.chicagofed.org/publications/workingpapers/wp2008_13.pdf (accessed
5/6/09); Jacobsen, Mark, 2008. "Evaluating U.S. Fuel Economy  Standards in a Model with Producer and
Household Heterogeneity," working paper, at http://econ.ucsd.edu/~m3jacobs/Jacobsen_CAFE.pdf (accessed
5/6/09).

49 Austin, David, and Terry Dinan, "Clearing the Air:  The Costs and Consequences of Higher CAFE Standards
and Increased Gasoline Taxes," Journal of Environmental Economics and Management 50 (2005): 562-582.

50IEA. 2007. "Mind the Gap: Quantifying Principal-Agent Problems in Energy Efficiency."  Paris, France:
International Energy Agency; Jaffe, Adam B., Richard G. Newell, and Robert N. Stavins (2001).  "Energy
Efficient Technologies and Climate Change Policies: Issues and Evidence." In Climate Change Economics and
Policy, Toman, Michael A., ed., Washington, D.C.: Resources for the Future, p. 171-181; Metcalf, Gilbert E., and
Kevin A. Hassett (1999). "Measuring the Energy Savings From Home Improvement Investments: Evidence
From Monthly Billing Data." The Review of Economics and Statistics 81(3):  516-528; Tietenberg, T. (2009).
"Reflections - Energy Efficiency Policy: Pipe Dream or Pipeline  to the Future?" Review of En vironmental
Economics and Policy. Vol.3, 2: 304-320.
                                               8-29

-------
Draft Regulatory Impact Analysis
51 Welch, David. "CAFE: A Prize for Making Gas-Guzzlers?"  Business Week April 16,2009,
http://www.businessweek.com/magazine/con ten t/09_17/b4128048030307. htm?chan=magazine+channel_what's
+next, accessed 7/7/09.

52 Berry, Steven, James Levinsohn, and Ariel Pakes (July 1995). "Automobile Prices in Market Equilibrium,"
Econometrica 63(4): 841-940.

  Greene, David L., Philip D. Patterson, Margaret Singh, and Jia Li, "Feebates, Rebates, and Gas-Guzzler
Taxes: A  Study of Incentives for Increased Fuel Economy," Energy Policy 33 (2005):  757-775.

54 Peer Review Report Summary: Estimating the Energy Security Benefits of Reduced U.S. Oil Imports, ICF,
Inc., September 2007.

55 Estimated reductions in imports of finished petroleum products and crude oil are 95% of 108 MMB in 2015,
353MMB  in 2020, 637MMB in 2030,and 740MMB in 2040.

56 Preliminary  Regulatory Impacts Analysis, April 2008. Based on a detailed analysis of differences in fuel
consumption, petroleum imports, and imports of refined petroleum products among the Reference Case, High
Economic Growth, and Low Economic Growth Scenarios presented in the Energy Information Administration's
Annual Energy Outlook 2007, NHTSA estimated that approximately 50 percent of the reduction in fuel
consumption is likely to be reflected in reduced U.S. imports of refined fuel, while the remaining 50 percent
would be expected to be reflected in reduced domestic fuel refining. Of this latter figure, 90 percent is
anticipated to reduce U.S. imports of crude petroleum for use as a refinery feedstock, while the remaining 10
percent is  expected to reduce U.S. domestic production of crude petroleum. Thus on balance, each gallon of fuel
saved is anticipated to reduc total U.S. imports of crude petroleum or refined fuel by 0.95 gallons..

57 See http://ostpxweb.dot.gov/policv/Data/VOT97guid.pdf and
http://ostpxweb.dot.gov/policy/Data/VOT revision l_2-ll-03.pdf

58  California Environmental Protection Agency, Air Resources Board. Draft Assessment of the Real-World
Impacts of Commingling California Phase 3 Reformulated Gasoline. August 2003

59 The 19.3 gallon average tank size is from EPA calculations conducted on the Volpe Model Market Data file
used in NHTSA'sModel Year 2011 CAFE Standards Final Rule.
60
  These benefits are included in the value of fuel savings reported in Tables VIII-5 through VIII-9.
61 These estimates were developed by FHWA for use in its 1997 Federal Highway Cost Allocation Study; see
http://www.fhwa.dot.gov/policy/hcas/fmal/index.htm (last accessed July 29, 2009).

62 See Federal Highway Administration, 1997 Federal Highway Cost Allocation Study,
http://www.fhwa.dot.gov/policy/hcas/final/index.htm, Tables V-22, V-23, and V-24 (last accessed July 27,
2009).

63 The Federal Highway Administration's estimates of these costs agree closely with some other recent estimates.
For example, recent published research conducted by Resources for the Future (RFF) estimates marginal
congestion and external accident costs for increased light-duty vehicle use in the U.S. to be 3.5 and 3.0 cents per
vehicle-mile in year-2002 dollars. See Ian W.H. Parry and Kenneth A. Small, "Does Britain or the U.S. Have
the Right Gasoline Tax?" Discussion Paper 02-12, Resources for the Future, 19 and Table 1 (March 2002).
Available at http://www.rff.org/rff/Documents/RFF-DP-02- 12.pdf (last accessed July 27, 2009).
                                                8-30

-------
                                               Small Business Flexibility Analysis

CHAPTER 9: Small Business Flexibility Analysis

       The Regulatory Flexibility Act, as amended by the Small Business Regulatory
Enforcement Fairness Act of 1996 (SBREFA), generally requires an agency to prepare a
regulatory flexibility analysis of any rule subject to notice-and-comment rulemaking
requirements under the Administrative Procedure Act or any other statute, unless the
agency certifies that the rule will not have a significant economic impact on a substantial
number of small entities. As a part of this analysis, an agency is directed to convene a
Small Business Advocacy Review Panel (SBAR Panel or 'the Panel'). During the Panel
process, we would gather information and recommendations from Small Entity
Representatives (SERs) on how to reduce the impact of the rule on small entities.

       The following provides an overview of small entities in the vehicle market.  Small
entities include small businesses, small organizations, and small governmental
jurisdictions. For the purposes of assessing the impacts  of the proposed rule on small
entities, a small entity is defined as: (1) a small business that meets the definition for
business based on the Small Business  Administration's (SBA) size standards (see Table
9-1); (2) a small governmental jurisdiction that is a government of a city, county, town,
school district or special district with a population of less than 50,000; and (3) a small
organization that is any not-for-profit enterprise which is independently owned and
operated and is not dominant in its field. Table 9-1 provides an overview of the primary
SBA small business categories potentially affected by this proposed regulation.

                Table 9-1: Primary Vehicle SBA Small Business Categories

Light-duty vehicle manufacturers
Vehicle importers
Alternative fuel vehicle converters
NAICSa Codes
336111
81111,811112
811198
Defined by SBA As a
small business if less than
or equal to :b
1,000 employees.
$7 million annual sales.
$7 million annual sales.
a. North American Industry Classification System
b. According to SBA's regulations (13 CFR 121), businesses with no more than the listed
number of employees or dollars in annual receipts are considered "small entities" for
RFA purposes.

       We compiled a list of vehicle manufacturers, independent commercial importers
(ICIs), and alternative  fuel converters that would be potentially affected by the proposed
rule from our 2008 model year certification databases.  These companies are already
certifying their vehicles for compliance with applicable EPA emissions standards (e.g.,
Tier 2). We then identified companies that appear to meet the definition of small
business provided in the table above. We were able to identify companies based on
certification information and previous rulemakings where we conducted Regulatory
Flexibility Analyses.

       Based on a preliminary assessment, EPA has identified a total of about 47 vehicle
entities, 33 of which are vehicle manufacturers. Of a total of 33 manufacturers, 2

                                       9-1

-------
Draft Regulatory Impact Analysis

manufacturers fit the SB A definition of a small entity. These businesses produce vehicles
for small niche markets, and all of these entities manufacture limited production, high
performance cars. Independent commercial importers (ICIs) are companies that hold a
Certificate (or Certificates) of Conformity permitting them to import nonconforming
vehicles and to modify these vehicles to meet U.S. emission standards. ICIs are not
required to meet the emission standards in effect when the vehicle is modified, but
instead they must meet the emission standards in effect when the vehicle was originally
produced (with an annual production cap of a total of 50 light-duty vehicles and trucks).
There are currently eight ICIs, all of which are small entities. Alternative fuel vehicle
converters are businesses that convert gasoline or diesel vehicles to operate on alternative
fuel (e.g., compressed natural gas), and converters must seek a certificate for all of their
vehicle models.  Model year 1993 and newer vehicles that are converted are required to
meet the  standards applicable at the time the vehicle was originally certified. Converters
serve a small niche market, and these businesses primarily convert vehicles to operate on
compressed natural gas (CNG) and liquefied petroleum gas (LPG), on a dedicated or dual
fuel basis. We identified  six alternative fuel converters in the light-duty vehicle market,
and three of these would qualify as small entities under SBA's definition.  Together, we
estimate that small entities comprise less than 0.1 percent of total annual vehicle sales
and deferring standards for them will have a negligible impact on the GHG emissions
reductions from the proposed standards.

       EPA has  not conducted a Regulatory Flexibility Analysis or a SBREFA SBAR
Panel for the proposed rule because we are proposing to certify that the rule would not
have a significant economic impact on a substantial number of small entities. EPA is
proposing to defer standards for manufacturers meeting SBA's definition of small
business as described in 13 CFR 121.201. EPA would instead consider appropriate GHG
standards for these entities as part of a future regulatory action. This includes small
entities in three distinct categories of businesses for light-duty vehicles: small volume
manufacturers, independent commercial importers (ICIs), and alternative fuel vehicle
converters. EPA has identified about 13 entities that fit the Small Business
Administration (SBA) criterion of a small business.  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 proposed deferment will have a negligible impact on the GHG emissions
reductions from the proposed standards.

       To ensure that EPA is aware of which companies would be deferred, EPA is
proposing 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. 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. Because such entities are not automatically
exempted from other EPA regulations for light-duty vehicles and light-duty trucks,  absent
such a declaration, EPA would assume that the entity was subject to the greenhouse gas
control requirements in this GHG proposal. The declaration would need to be submitted
at time of vehicle emissions certification under the EPA Tier 2 program. EPA expects
that the additional paperwork burden associated with completing and submitting a small
                                       9-2

-------
                                               Small Business Flexibility Analysis

entity declaration to gain deferral from the proposed GHG standards would be negligible
and easily done in the context of other routine submittals to EPA. However, EPA has
accounted for this cost with a nominal estimate included in the Information Collection
Request completed under the Paperwork Reduction Act. Additional information can be
found in the Paperwork Reduction Act discussion in section III.I.2.  Based on this, EPA
is proposing to certify that the rule would not have a significant economic impact on a
substantial number of small entities.
                                       9-3

-------