Final Rulemaking to Establish Light-Duty
Vehicle Greenhouse Gas Emission
Standards and Corporate Average Fuel
Economy Standards
Regulatory Impact Analysis
United States
Environmental Protection
Agency
-------
Final Rulemaking to Establish Light-Duty
Vehicle Greenhouse Gas Emission
Standards and Corporate Average Fuel
Economy Standards
Regulatory Impact Analysis
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
SER&
United States
Environmental Protection
Agency
EPA-420-R-10-009
April 2010
-------
EPA FINAL RIA TABLE OF CONTENTS
Executive Summary ES-1
CHAPTER 1: TECHNOLOGY PACKAGES, COST AND EFFECTIVENESS
1.1 Overview of Technology 1-1
1.2 Technology Cost and Effectiveness 1-4
1.3 Package Cost and Effectiveness 1-11
1.3.1 Explanation of Technology Packages 1-11
1.3.2 Technology Package Costs & Effectiveness 1-13
1.4 EPA's Lumped Parameter Approach for Determining Effectiveness Synergies 1-25
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 1-32
1.4.5 Technologies Considered by EPA and Ricardo in the Vehicle Simulation 1-33
1.4.6 Choice of Technology Packages 1-35
1.4.7 Simulation Results 1-37
1.5 Comparison of Lumped-Parameter Results to Modeling Results 1-38
1.6 Using the Lumped-Parameter Technique to Determine Synergies in a Technology
Application Flowpath 1-40
CHAPTER 2: AIR CONDITIONING
2.1 Overview of Air Conditioning Impacts and Technologies 2-1
2.2 Air Conditioner Leakage 2-3
2.2.1 Impacts of Refrigerant Leakage on Greenhouse Gas Emissions 2-3
2.2.2 A/C Leakage Credit 2-5
2.2.3 Technologies That Reduce Refrigerant Leakage and their Effectiveness 2-12
2.2.4 Technical Feasibility of Leakage-Reducing Technologies 2-16
2.2.5 Leakage Controls in A/C Systems 2-16
2.2.6 Other Benefits of improving A/C Leakage Performance 2-19
2.3 CO2 Emissions due to Air Conditioners 2-20
2.3.1 Impact of Air Conditioning Use on Fuel Consumption and CO2 Emissions 2-20
2.3.2 Technologies That Improve Efficiency of Air Conditioning and Their
Effectiveness 2-30
2.3.3 Technical Feasibility of Efficiency-Improving Technologies 2-35
2.3.4 A/C Efficiency Credits 2-35
2.4 Costs of A/C reducing technologies 2-43
2.5 Air Conditioning Credit Summary 2-45
CHAPTER 3: TECHNICAL BASIS OF THE STANDARDS
3.1 Technical Basis of the Standards 3-1
3.1.1 Summary 3-1
3.1.2 Overview of Equivalency Calculation 3-1
11
-------
3.2 Analysis of Footprint Approach for Establishing Individual Company Standards 3-6
3.2.1 "Footprint" as a Vehicle Attribute 3-7
3.2.2 Alternative Attributes 3-12
3.2.3 EPA Selection of the Footprint Attribute 3-19
CHAPTER 4: RESULTS OF FINAL AND ALTERNATIVE STANDARDS
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 Manufacturer-Specific Standards and Achieved CO2 Levels 4-16
4.6 Per Vehicle Costs 2012-2016 4-17
4.7 Technology Penetration 4-19
4.8 Alternative Program Stringencies 4-26
4.9 Assessment of Manufacturer Differences 4-36
CHAPTER 5: EMISSIONS IMPACTS
5.1 Overview 5-1
5.2 Introduction 5-4
5.2.1 Scope of Analysis 5-4
5.2.2 Downstream Contributions 5-4
5.2.3 Upstream Contributions 5-5
5.2.4 Global Warming Potentials 5-5
5.3 Program Analysis and Modeling Methods 5-6
5.3.1 Models Used 5-6
5.3.2 Description of Scenarios 5-7
5.3.3 Calculation of Downstream Emissions 5-16
5.3.4 Calculation of Upstream Emissions 5-30
5.4 Greenhouse Gas Emission Inventory 5-30
5.5 Non-Greenhouse Gas Emission Inventory 5-31
5.5.1 Downstream Impacts of Program on Non-GHG Emissions 5-33
5.5.2 Upstream Impacts of Program on Non-GHG Emissions 5-34
5.5.3 Total non-GHG Program Impact 5-35
5.6 Model Year Lifetime Analyses 5-36
5.6.1 Methodology 5-36
5.6.2 Results 5-38
5.7 Alternative 4% and 6% Scenarios 5-40
5.7.1 4% Scenario 5-40
5.7.2 6% Scenario 4-42
5.8 Inventories Used for Air Quality Analyses 5-44
5.8.1 Upstream Emissions 5-44
5.A Appendix to Chapter 5: Details of the TLAAS Impacts
Analysis 5-49
111
-------
5.A.1 Introduction and Summary 5-49
5.A.2 Factors Determining the Impact of the TLAAS 5-50
5.A.3 Bounding Analysis of TLAAS Impact 5-51
5.A.4 Approach used for Estimating TLAAS Impact 5-51
5.B Appendix to Chapter 5: Impacts of Advanced Technology Vehicle Incentives for
Electric Vehicles 5-54
5.B.1 Introduction and Summary 5-54
5.B.2 Assumptions behind the Analysis 5-54
5.B.3 Inputs 5-55
5.B.4 Computation 5-56
CHAPTER 6: VEHICLE PROGRAM COSTS INCLUDING FUEL CONSUMPTION
IMPACTS
6.1 Vehicle Program Costs 6-2
6.1.1 Vehicle Compliance Costs on a Per-Vehicle Basis 6-2
6.1.2 Vehicle Compliance Costs on a Per-Year Basis 6-10
6.2 Cost per Ton of Emissions Reduced 6-14
6.3 Fuel Consumption Impacts 6-14
6.4 Vehicle Program Cost Summary 6-18
CHAPTER 7: ENVIRONMENTAL AND HEALTH IMPACTS
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 7-17
7.2 Non-GHG Air Quality Impacts 7-30
7.2.1 Air Quality Modeling Methodology 7-31
7.2.2 Air Quality Modeling Results 7-44
7.3 Quantified and Monetized Non-GHG Health and Environmental Impacts 7-86
7.3.1 Quantified and Monetized Non-GHG Human Health Benefits of the 2030 Calendar
Year (CY) Analysis 7-87
7.3.2 PM-related Monetized Benefits of the Model Year (MY) Analysis 7-117
7.4 Changes in Global Mean Temperature and Sea Level Rise Associated with the Rule's
GHG Emissions Reductions 7-122
7.4.1 Introduction 7-122
moran7.4.2 Estimated Projected Reductions in Atmospheric CO2 Concentrations, Global
Mean Surface Temperatures and Sea Level Rise 7-122
7.4.3 Summary 7-126
7.5 SCC and GHG Benefits 7-127
7.6 Weight Reduction and Vehicle Safety 7-135
7.6.1 What did EPA say in the NPRM and in the Draft RIA with regard to Potential Safety
Effects? 7-136
7.6.2 What Public Comments did EPA Receive in Regard to its Safety Discussion and
what is EPA's Response? 7-142
7.6.3 NHTSA's 2010 Study of Accident Fatalities by Vehicle Size and Weight 7-146
7.6.4 Suggested Next Steps To Increase Our Understanding of the Effects of Vehicle
Size and Weight on Fatalities 7-148
IV
-------
CHAPTER 8: OTHER ECONOMIC AND SOCIAL IMPACTS
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 ....8-13
8.2 Energy Security Impacts 8-16
8.3 Other Impacts 8-18
8.3.1 Reduced Refueling Time 8-18
8.3.2 Value of Additional Driving 8-19
8.3.3 Noise, Congestion, and Accidents 8-19
8.3.4 Summary of Other Impacts 8-21
8.4 Summary of Costs and Benefits 8-23
CHAPTER 9: SMALL BUSINESS FLEXIBILITY 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
-------
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-cylinder engine in a "V" configuration
V8: 8-cylinder engine in a "V" configuration
WGII: Working group II
vn
-------
Executive Summary
The Environmental Protection Agency (EPA) and the National Highway Traffic
Safety Administration (NHTSA) are issuing a joint final rulemaking to establish new
standards for light-duty highway vehicles that will reduce greenhouse gas emissions (GHG)
and improve fuel economy. The joint 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 finalizing greenhouse gas emissions standards under the Clean Air Act, and NHTSA is
finalizing Corporate Average Fuel Economy standards under the Energy Policy and
Conservation Act (EPCA), as amended. These standards apply to passenger cars, light-duty
trucks, and medium-duty passenger vehicles, covering model years (MY) 2012 through 2016.
The standards will require these vehicles to meet an estimated combined average emissions
level of 250 grams of CO2 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 National Program will result in approximately 960
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.
Mobile sources are significant contributors to air pollutant emissions (both GHG and
non-GHG) across the country, internationally, and into the future. The Agency has
determined that these emissions cause or contribute to air pollution which may reasonably be
anticipated to endanger public health or welfare, and is therefore establishing standards to
control these emissions as required by section 202 (a) of the Clean Air Act.1 The health- and
environmentally-related effects associated with these emissions are a classic example of an
externality-related market failure. An externality occurs when one party's actions impose
uncompensated costs on another party. EPA's final rule will deliver additional environmental
and energy benefits, as well as cost savings, on a nationwide basis that would likely not be
available if the rule were not in place.
Table 1 shows EPA's estimated lifetime discounted cost, benefits and net benefits for
all vehicles projected to be sold in model years 2012-2016. It is important to note that there is
significant overlap in costs and benefits for NHTSA's CAFE program and EPA's GHG
program and therefore combined program costs and benefits are not a sum of the individual
programs.
1 "Technical Support Document for Endangerment and Cause or Contribute Findings for Greenhouse Gases
Under Section 202(a) of the Clean Air Act" Docket: EPA-HQ-OAR-2009-0472-11292,
http://epa.gov/climatechange/endangerment.html. See also State of Massachusetts v. EPA, 549 U.S. 497, 533
("If EPA makes a finding of endangerment, the Clean Air Act requires the agency to regulate emissions of the
deleterious pollutant from new motor vehicles").
ES-1
-------
Table 1 EPA's Estimated 2012-2016 Model Year Lifetime Discounted Costs,
Benefits, and Net Benefits assuming the $21/ton SCC
(Millions of 2007 dollars)
3% Discount Rate
Costs
Benefits
Net Benefits
7% Discount Rate
Costs
Benefits
Net Benefits
$51,500
$240,200
$188,700
$51,500
$191,700
$140,200
a As noted in Section III.H, SCC increases over time. The $21/ton value applies to
2010 CO2 emissions and grows larger over time.
b Although EPA estimated the benefits associated with four different values of a one
ton GHG reduction ($5, $21, $35, $65), for the purposes of this overview presentation
of estimated costs and benefits EPA is showing the benefits associated with the
marginal value deemed to be central by the interagency working group on this topic:
$21 per ton of CO2e, in 2007 dollars and 2010 emissions. The $21/ton value applies
to 2010 CO2 emissions and grows over time.
c Note that net present value of reduced GHG emissions is calculated differently than
other benefits. The same discount rate used to discount the value of damages from
future emissions (SCC at 5, 3, and 2.5 percent) is used to calculate net present value of
SCC for internal consistency. Refer to Section III.H for more detail.
d Monetized GHG benefits exclude the value of reductions in non-CO2 GHG
emissions (HFC, CH4 and N2O) expected under this final rule. Although EPA has not
monetized the benefits of reductions in these non-CO2 emissions, the value of these
reductions should not be interpreted as zero. Rather, the reductions in non-CO2 GHGs
will contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC
TSD notes the difference between the social cost of non-CO2 emissions and CO2
emissions, and specifies a goal to develop methods to value non-CO2 emissions in
future analyses.
This Regulatory Impact Analysis (RIA) contains supporting documentation to the
EPA rulemaking. NHTS A has prepared their own RIA in support of their rulemaking (this
can be found in NHTS A's docket for the rulemaking, NHTSA-2009-0059). While the two
rulemakings are similar, there are also differences in the analyses that require separate
discussion. This is largely because EPA and NHTS A act under different statutes. EPA's
authority comes under the Clean Air Act, and NHTS A'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 joint Technical Support Document (joint
TSD can be found in EPA's docket EPA-HQ-OAR-2009-0472). Therefore, this RIA should
be viewed as a companion document to the Joint TSD and the two documents together
provide the details of EPA's technical analysis in support of its rulemaking.
While NHTSA and EPA each modeled their respective regulatory programs under the
National Program, the analyses were generally consistent and featured similar parameters.
EPA did not conduct an overall uncertainty analysis of the impacts associated with its
regulatory program, though it did conduct uncertainty and sensitivity analyses of individual
components of the analysis (e.g., uncertainty ranges associated with quantified and monetized
ES-2
-------
non-GHG health impacts). NHTSA, however, conducted a Monte Carlo simulation of the
uncertainty associated with its regulatory program. The focus of the simulation model was
variation around the chosen uncertainty parameters and their resulting impact on the key
output parameters, fuel savings, and net benefits. Among other parameters, NHTSA varied
technology costs, technology effectiveness, fuel prices, the values of oil consumption
externalities and the rebound effect. Because of the similarities between the two analyses,
EPA references NHTSA RIA Chapter XII as indicative of the relative magnitude, uncertainty
and sensitivities of parameters of the cost/benefit analysis.
This document contains the following;
Chapter 1: Technology Packages, Cost and Effectiveness. The details of the vehicle
technology packages used as inputs to EPA's Optimization Model for Emissions of
Greenhouse gases from Automobiles (OMEGA) are presented. These vehicle packages
represent potential ways of meeting the CO2 stringency established by this rule and are the
basis of the technology costs and effectiveness analyses discussed in Chapter 3 of the Joint
TSD. This chapter also contains details on the lumped parameter model, which is a major
part of EPA's determination of the effectiveness of these packages.
Chapter 2: Air Conditioning. EPA's unique air conditioning (A/C) program is
discussed. Details for this chapter include the A/C credit program and the related technology
costs and effectiveness associated with new A/C systems. The A/C credit program allows
manufacturers to earn credit for both direct and indirect CO2eq emissions. Direct emission
credits are earned through reducing refrigerant leakage (as the current refrigerant, R-134a, has
a very high global warming potential) from the A/C system. The amount of direct emission, or
leakage, credit that a manufacturer can earn is determined by using a design-based method to
calculate the yearly refrigerant leakage from a vehicle's A/C system. This leakage value is
then used to calculate a "grams-per-mile" credit, with allowances made for the global
warming potential of the refrigerant. Indirect emission credits are earned through improving
the efficiency of the A/C system, which reduces the amount of power required to operate the
A/C system as well as the amount of CO2 emitted by the vehicle. The amount of indirect
emission credit is determined by using a menu-based approach, where the inclusion of
specific efficiency-improving components or design elements into a vehicle's A/C system
results in an assigned credit value.
Chapter 3: Technical Basis of the Standards. This chapter contains two subchapters.
In the first, EPA evaluates 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 CO2 performance standards for model year 2016 which would result in the
new vehicle fleet in the State of California having CO2 performance equal to the performance
from the California Pavley 1 standards. This technical analysis is documented in this sub-
chapter of the RIA. In the second subchapter, EPA discusses an analysis of the "footprint"
approach EPA is using for establishing standards.
Chapter 4: Results of Final and Alternative Standards. A conceptual overview of
EPA's OMEGA model and technology cost results for the program and alternative standards
ES-3
-------
considered. For each manufacturer, EPA estimates the following achieved fleetwide CO2
levels and technology costs over the reference case, see Table 2.
Table 2 Fleetwide Costs in 2016
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Total
Fleetwide Cost
$ 1,453
$ 1,329
$ 1,231
$ 899
$ 1,219
$ 575
$ 745
$ 984
$ 594
$ 808
$ 1,343
$ 978
$ 810
$ 1,257
$ 937
$ 455
$ 1,694
$ 948
Key results of the alternative stringencies analyzed are presented in Table 3:
Table 3 Key Results Per Reduction Scenarios of 4% and 6% Per Year
Reduction Scenarios
4% per year
6% per year
Industry
achieved CO2
level (g/mi)
256.9
236.1
Average
industry cost
per vehicle
$883
$1,343
ES-4
-------
Chapter 5: Emissions Impacts. This chapter contains the greenhouse gas and non-
greenhouse gas emission impacts of this rule and includes the impact of credits.
Greenhouse Emission Impacts of EPA's Rulemaking
Table 4 shows reductions estimated from EPA's GHG standards. The analyses
assume a pre-control case of MY 2011 CAFE standards continuing indefinitely beyond 2011,
and a post-control case in which MY 2016 standards continue indefinitely beyond 2016.
Including the reductions from upstream emissions (fuel production and transport),
total reductions are estimated to reach 307 MMTCC^eq (million metric tons of COi
equivalent emissions) annually by 2030 (equivalent to a 21 percent reduction in U.S. car and
light truck emissions as compared to the reference scenario), and grow to over 500
MMTCCheq in 2050 as cleaner vehicles continue to come into the fleet (equivalent to a 23
percent reduction in U.S. car and light truck emissions relative to the control case that year).
Table 4. Projected Net GHG Reductions (MMT CO2 Equivalent per year)
Calendar Year:
Net Reduction*
Net CO2
Net other GHG
Downstream Reduction
CO2 (excluding A/C)
A/C- indirect CO '2
A/C - direct HFCs
CH4 (rebound effect)
N2O (rebound effect)
Upstream Reduction
CO2
CH4
N2O
2020
156.4
139.1
17.3
125.2
101.2
10.6
13.3
0.0
0.0
31.2
27.2
3.9
0.1
2030
307.0
273.3
33.7
245.7
799.5
20.2
26.0
0.0
-0.1
61.3
53.5
7.6
0.3
2040
401.5
360.4
41.1
320.7
263.2
26.5
30.9
0.0
-0.1
80.8
70.6
10.0
0.3
2050
505.9
458.7
47.2
403.0
335.1
33.8
34.2
0.0
-0.1
102.9
89.9
12.7
0.4
* includes impacts of 10% VMT rebound rate
Impacts of EPA's Rulemaking on Emissions of Criteria and Toxic Pollutants
The results of EPA's analyses on the impacts of the program on annual criteria
emissions are listed in Table 5. For all criteria pollutants the overall impact of the program
will be relatively small compared to total U.S. inventories across all sectors. In 2030, EPA
estimates the program will reduce total NOx, PM and SOx inventories by 0.1 to 0.8 percent
and reduce the VOC inventory by 1.0 percent, while increasing the total national CO
inventory by 0.6 percent.
ES-5
-------
EPA estimates that the GHG program will result in small changes for toxic emissions
compared to total U.S. inventories across all sectors, as listed in Table 6. In 2030, EPA
estimates the program will reduce total benzene and 1,3 butadiene emissions by 0.1 to 0.3
percent. Total acrolein and formaldehyde emissions will increase by 0.1 percent.
Acetaldehyde emissions will increase by 2.2 percent.
Table 5 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
-60,187
-0.51%
3,992
0.01%
-5,881
-0.02
-2,398
-0.03%
-13,832
-0.41%
2030
-115,542
-1.01%
170,675
0.56%
-21,763
-0.07%
-4,564
-0.05%
-27,443
-0.82%
Upstream Impacts
2020
-64,506
-0.55%
-6,165
-0.02%
-19,291
-0.06%
-2,629
-0.03%
-11,804
-0.35%
2030
-126,749
-1.11%
-12,113
-0.04%
-37,905
-0.12%
-5,165
-0.06%
-23,194
-0.69%
Downstream Impacts
2020
4,318
0.04%
10,156
0.01%
13,410
0.04%
231.0
0.00%
-2,027
-0.06%
2030
11,207
0.01%
182,788
0.6%
16,143
0.05%
602.3
0.01%
-4,249
-0.13%
Table 6. 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
-95
-0.38%
760
2.26%
1
0.01%
-890
-0.48%
-49
-0.06%
2030
-21
-0.10%
668
2.18%
5
0.07%
-523
-0.29%
15
0.02%
Upstream Impacts
2020
-1.5
-0.01%
-6.8
-0.02%
-0.9
-0.01%
-139.6
-0.08%
-51.4
-0.06%
2030
-3.0
-0.01%
-13.4
-0.04%
-1.8
-0.03%
-274.3
-0.15%
-101.0
-0.12%
Downstream Impacts
2020
-93.6
-0.37%
766.9
2.28%
1.7
0.03%
-750.0
-0.40%
2.1
0.00%
2030
-18.1
-0.09%
681.5
2.22%
6.5
0.10%
-248.3
-0.14%
116.3
0.14%
ES-6
-------
Chapter 6: Vehicle Program Costs Including Fuel Consumption Impacts. The
program costs and fuel savings associated with EPA's rulemaking. In Chapter 6, we present
briefly some of the outputs of the OMEGA model (costs per vehicle) and how we use those
outputs to estimate the annual costs (and fuel savings) of the program through 2050. We also
present our cost per ton analysis showing the cost incurred for each ton of GHG reduced by
the program.
Chapter 7: Environmental and Health Impacts. This Chapter provides details on the
non-GHG health and environmental impacts associated with criteria pollutants and air toxics.
We also present the results of our non-GHG air quality modeling analysis and the quantified
and monetized estimates of PM2.5- and ozone-related health impacts. Our air quality modeling
indicates that the final standards have relatively little impact on ambient concentrations of
modeled PM2.5, ozone, and air toxics (See Chapter 7.2.). The criteria pollutant-related
benefits of the rule are associated with small reductions in PM2.5.
As described in Chapter 7.5, EPA used four new estimates of the dollar value of
marginal reductions in CO2 emissions—known as the social cost of carbon—to calculate total
monetized CO2 benefits. Specifically, total monetized benefits in each year are calculated by
multiplying the SCC by the reductions in CO2 for that year. EPA used four different SCC
values to generate different estimates of total CO2 benefits and capture some of the
uncertainties involved in regulatory impact analysis. The central value is the average SCC
across models at the 3 percent discount rate. For purposes of capturing the uncertainties
involved in regulatory impact analysis, we emphasize the importance and value of considering
the full range. Chapter 7 also presents an analysis of the CO2 benefits over the model year
lifetimes of the 2012 through 2016 model year vehicles.
Chapter 7.6 also includes additional information about EPA's mass reduction and
safety analysis.
Chapter 8: Other Economic and Social Impacts. This chapter provides a description
of other economic and social impacts associated with the rule, including vehicle sales impacts,
consumer vehicle choice, energy security, and other economic impacts associated with
reduced refueling, the value of increased driving, and the cost associated with additional
noise, congestion and accidents.
Vehicle Sales Impacts: Our analysis predicts vehicle sales increasing as a result of the
rule. 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 at least five
years of 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).
Consumer Choice Impacts: Consumer vehicle choice models could in principle be
used to examine the effects of this rule on the mix of vehicles sold. In practice, however,
ES-7
-------
EPA finds that the state of the art of these models is not yet settled. The models show great
variation, and there has been very little comparative assessment of them. 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 rule.
Payback Period on New Vehicle Purchases: We also conducted what we call our
"payback analysis" which looks at how quickly the improved fuel efficiency of new vehicles
provides savings to buyers despite the vehicles having new technology (and new costs). The
consumer payback analysis shows that fuel savings will outweigh up-front costs within three
years for people purchasing new vehicles with cash. For those purchasing new vehicles with
a typical five-year car note, the fuel savings will outweigh increased costs in the first month of
ownership.
Energy Security Impacts: A reduction of U.S. petroleum imports reduces both
financial and strategic risks associated with a potential disruption in supply or a spike in cost
of a particular energy source. This reduction in risk is a measure of improved U.S. energy
security. Based on these estimates of fuel savings, over the lifetimes of model years 2012-
2016, we estimate the discounted energy security impacts at $10.1 billion dollars, in 2007
dollars, assuming a 3 percent discount rate, and $8.0 billion dollars, assuming a 7 percent
discount rate.
Other Impacts: There are other impacts associated with the 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 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 standards, but they are nevertheless important to include.
Chapter 8 also presents a summary of the total costs, total benefits, and net benefits
expected under the final rule. Table 7 presents these economics impacts.
Chapter 9: Small Business Flexibility Analysis. EPA's analysis of the small business
impacts due to EPA's rulemaking.
ES-8
-------
Table 7 Economic Impacts of the Light-Duty GHG Rule
Vehicle Costs
Fuel Savings6
Reduced Refueling
2020 | 2030
$15,600 $15,800
$2';4"66 $4,800
2040 |
$17,400
$673(X)'
Benefits from Reduced CO2 Emissions at each assumed SCC valuec>cl'e
AvgSCCat5% $900 $2,700 $4,600
""Xvg"SCC"at3% $37766 $87900 ^OOO
Aygscc^
2050
$19,000
$8,666
$7,200
$36,666
$627666
NPV, 3%a |
$345,900
$87^9"66
$34,500
$Y76j66
$2997666
$5387566
NPV, 7%a
$191,900
$34,500
$i76j00
$2997666
Other Impacts
Criteria Pollutant $1,200- $1,200- $1,200-
Benefits^ B $1,300 $1,300 $1,300 $2.1,000 $14,000
Energy Security Impacts
(price shock) $2,200 $4,500 $6,000 $7,600 $81,900 $36,900
Accidents, Noise,
Congestion -$2,300 -$4,600 -$6,100 -$7,800 -$84,800 -$38,600
Quantified Net Benefits at each
AvgSCCajt5%
assumed SCC valuec'd'e
$27,500 $81,500
$30^300 $87jo6
$32^406 $92^06
$127,000
$136,466
$Y437466
$186,900
$209j66
$1,511,700
$643,100
$9087266
a Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used to
calculate net present value of SCC for internal consistency. Refer to Chapter 7 for more detail.
b Calculated using pre-tax fuel prices.
c Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O)
expected under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2
emissions, the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2
GHGs will contribute to this rule's climate benefits, as explained in Section III.F.2 of the preamble. The SCC
Technical Support Document (TSD) notes the difference between the social cost of non-CO2 emissions and CO2
emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses.
d Section III.H.6 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates
range as follows: for Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $35-$65;andfor95thpercentileSCCat3%: $65-$136. Section III.H.6 also presents these SCC
estimates.
e Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used to
calculate net present value of SCC for internal consistency. Refer to SCC TSD for more detail.
f Note that "B" indicates unquantified criteria pollutant benefits in the year 2020. For the final rule, we only
modeled the rule's PM2.5- and ozone-related impacts in the calendar year 2030. For the purposes of estimating
a stream of future-year criteria pollutant benefits, we assume that the benefits out to 2050 are equal to, and no
less than, those modeled in 2030 as reflected by the stream of estimated future emission reductions. The NPV of
criteria pollutant-related benefits should therefore be considered a conservative estimate of the potential benefits
associated with the final rule.
g The benefits presented in this table include an estimate of PM-related premature mortality derived from Laden
et al., 2006, and the ozone-related premature mortality estimate derived from Bell et al., 2004. If the benefit
estimates were based on the ACS study of PM-related premature mortality (Pope et al., 2002) and the Levy et al.,
2005 study of ozone-related premature mortality, the values would be as much as 70% smaller.
h The calendar year benefits presented in this table assume either a 3% discount rate in the valuation of PM-
related premature mortality ($1,300 million) or a 7% discount rate ($1,200 million) to account for a twenty-year
segmented cessation lag. Note that the benefits estimated using a 3% discount rate were used to calculate the
ES-9
-------
NPV using a 3% discount rate and the benefits estimated using a 7% discount rate were used to calculate the
NPV using a 7% discount rate. For benefits totals presented at each calendar year, we used the mid-point of the
criteria pollutant benefits range ($1,250).
1 Note that the co-pollutant impacts presented here do not include the full complement of endpoints that, if
quantified and monetized, would change the total monetized estimate of impacts. The full complement of
human health and welfare effects associated with PM and ozone remain unquantified because of current
limitations in methods or available data. We have not quantified a number of known or suspected health effects
linked with ozone and PM for which appropriate health impact functions are not available or which do not
provide easily interpretable outcomes (e.g., changes in heart rate variability). Additionally, we are unable to
quantify a number of known welfare effects, including reduced acid and particulate deposition damage to
cultural monuments and other materials, and environmental benefits due to reductions of impacts of
eutrophication in coastal areas.
ES-10
-------
Technology Packages, Cost and Effectiveness
CHAPTER 1: Technology Packages, Cost and Effectiveness
1.1 Overview of Technology
The final 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 particulate 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
-------
Regulatory Impact Analysis
to address COi. Instead, the focus of the CCh 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 COi emissions by either making
more efficient use of the energy that is produced through combustion of the fuel or by
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 final 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 cannot reasonably redesign
any given vehicle every year or even every other year, let alone redesign all of their vehicles
every year or every other year. 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.
1-2
-------
Technology Packages, Cost and Effectiveness
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 typically involves significant engineering, development, manufacturing, and marketing
resources to create a new product with multiple new features. In order to leverage this
significant upfront investment, manufacturers plan vehicle redesigns with several model
years' of production in mind. That said, 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.
Given that the regulatory timeframe of the GHG program is five years (2012 through
2016), and given EPA's belief that full line manufacturers (i.e., those making small cars
through large cars, minivans, small trucks and large trucks) cannot redesign, on average, their
entire product line more than once during that timeframe, a five year redesign cycle has been
used in our final rule. This same redesign cycle was used in the proposal. This means that the
analysis assumes that each vehicle platform in the US fleet can undergo at least one full
redesign during our regulatory timeframe. While some may undergo more than one, the
analysis assumes that the extra redesign comes at the expense of another vehicle that would,
in effect, undergo no redesign during the regulatory timeframe.
Commenters were generally supportive of the use of a five year redesign cycle.
However, at least one commenter argued that shorter redesign cycles are possible and that the
final GHG standards are too low in light of the ability of manufacturers to conduct redesigns
at a faster pace. EPA's response on both sides of this issue can be found in the Response to
Comments document (see issue 3.1).
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 CC>2, and as discussed
below, to reduce emissions of HFCs from the air conditioner.
1-3
-------
Regulatory Impact Analysis
In determining the projected technology needed to meet the standards, and the cost of
those technologies, EPA is using 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 rule is
commercially available and already being employed to a limited extent across the fleet,
although far greater penetration of these technologies into the fleet is projected as a result of
the final rule. The vast majority of the emission reductions which will result from the rule
would result from the increased use of these technologies. EPA also believes the rule would
encourage the development and limited use of more advanced technologies, such as PHEVs
and EVs, and is structuring the rule to encourage these technologies' use.
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.
1.2 Technology Cost and Effectiveness
EPA collected information on the cost and effectiveness of COi 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
NHTSA 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.
Since publication of the 2011 CAFE FRM and EPA's 2008 Staff Technical Report,
EPA began a contracted study 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
1-4
-------
Technology Packages, Cost and Effectiveness
very detailed estimates of the costs associated with manufacturing them. Also, cost and
effectiveness estimates were adjusted as a result of further meetings between EPA and
NHTSA 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 CO2 emission control, such as papers published by the Society of
Automotive Engineers and the American Society of Mechanical Engineers. The results of
these efforts and early results of the FEV contracted study were used in the proposal for this
rule.
Since the proposal, EPA has carefully examined all information on technology cost
and effectiveness received during the comment period. Importantly, the FEV contracted
study has progressed and provides many more new cost estimates that have been incorporated
in the final analysis. As a result, while some FEV teardown costs were used in the proposal,
we have expanded our use of FEV costs for the final rule using new information available to
us shortly after the proposal. For more detail on our technology cost estimates and how they
have changed since the proposal refer to Chapter 3 of the joint TSD and, specifically, section
3.3.2.2 of the joint TSD for how costs have changed.
EPA reviewed all this information in order to develop the best estimates of the cost
and effectiveness of COi 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 effectiveness of COi reducing
technologies can be found in Chapter 3 of the joint TSD, along with a more detailed
description of the comprehensive technical evaluation underlying the estimates.
1-5
-------
Regulatory Impact Analysis
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
c«
G
°5b
W
>
ffi
O
Turbo w/o downsize
Downsize w/o turbo
Turbo with downsize
Low friction lubricants
Engine friction reduction
VVT - intake cam
phasing
VVT - coupled cam
phasing
VVT - dual cam phasing
Cylinder deactivation
Discrete VVLT
Continuous VVLT
Cylinder deactivation
VVT - coupled cam
phasing
Discrete VVLT
Continuous VVLT
(includes conversion to
Overhead Cam)
Camless valvetrain
(electromagnetic)
GDI - stoichiometric 14
GDI - stoichiometric V6
GDI - stoichiometric V8
GDI - lean burn
Turbocharge (single)
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
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 14
Base V6
Base V8
GDI - stoich
Base 14
Base V6
Base engine
V6 DOHC
V6 SOHC
V6OHV
14 DOHC (larger)
14 DOHC
V8 DOHC
V8 SOHC 2V
V8 SOHC 3V
V80HV
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
Vehicle Class
Small Car
$3
$50
$40
$40
$73
n/a
$125
$245
n/a
$40
$141
$497
$501
$209
n/a
n/a
$623
$397
n/a
$666
n/a
n/a
n/a
-$67
-$116
n/a
n/a
n/a
n/a
n/a
n/a
$771
$391
$349
n/a
n/a
Large Car
$3
$75
$80
$80
$157
$150
$181
$449
$150
$40
$204
$1,048
$501
209
$301
$346
$623
n/a
$420
$666
-$384
-$177
$265
-$67
n/a
-$188
$59
-$17
$310
$149
$323
$771
n/a
n/a
$592
$816
Minivan
$3
$75
$80
$80
$157
$150
$181
$449
$150
$40
$204
$1,048
$501
$209
$301
$346
$623
$397
$420
$666
-$384
-$177
$265
-$67
n/a
-$188
$59
-$17
$310
$149
$323
$771
$391
n/a
$592
$816
Small Truck
$3
$75
$80
$80
$157
$150
$181
$449
$150
$40
$204
$1,048
$501
209
$301
n/a
$623
n/a
$420
$666
-$384
-$177
$265
-$67
n/a
-$188
$59
-$17
$310
$149
$323
$771
n/a
n/a
$592
$816
Large Truck
$3
$100
$80
$80
$157
$169
$259
$489
$169
$40
$291
$1,146
$501
209
$301
$346
$623
n/a
$420
$666
-$384
-$177
$265
-$67
n/a
-$188
$59
-$17
$310
$149
$323
$771
n/a
n/a
$592
$816
1-6
-------
Technology Packages, Cost and Effectiveness
Downsize to V6 DOHC
& add twin turbo
Downsize to V6 DOHC
& 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
V8 SOHC 3V w/o
turbo
V8 OHV w/o turbo
V6 SOHC
V6OHV
V8 SOHC 2V
V8 SOHC 3V
V80HV
GDI - stolen
Base gasoline engine
Base gasoline engine
n/a
n/a
n/a
n/a
n/a
n/a
n/a
$253
$1,877
$736
$1,099
$258
$464
$292
$213
$509
$375
$2,655
$736
$1,099
$258
$464
$292
$213
$509
$375
$2,164
$736
$1,099
$258
$464
$292
$213
$509
$375
$2,164
$736
$1,099
$258
$464
$292
$213
$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
$99
-$52
-$7
$79
$192
Large
Car
$28
$25
$90
$99
-$52
-$7
$79
$224
Minivan
$28
$25
$90
$99
-$52
-$7
$79
$224
Small
Truck
$28
$25
$90
$99
-$52
-$7
$79
n/a
Large
Truck
$28
$25
$90
$99
-$52
-$7
$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
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
IMA/ISA/BSG
hybrid
Power-split
hybrid
Base engine &
trans
Vehicle Class
Small Car
$351
$2,854
$4,232
$3,967
$6,922
$5,423
$27,628
Large Car
$398
$3,612
$5,469
$5,377
$9,519
$7,431
n/a
Minivan
$398
$3,627
$5,451
$5,378
$9,598
$7,351
n/a
Small Truck
$398
$3,423
$4,943
$4,856
$9,083
$7,128
n/a
Large Truck
$437
$4,431
$7,236
$7,210
$12,467
$9,643
n/a
1-7
-------
Regulatory Impact Analysis
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
1 2 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 CC>2 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 joint TSD and each of these estimates is discussed in more
detail there.
1-8
-------
Technology Packages, Cost and Effectiveness
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
VVT - intake cam phasing
VVT - 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
1-9
-------
Regulatory Impact Analysis
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)
Absolute CO2 Reduction (% from baseline vehicle)
Small
Car
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
Large
Car
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
Minivan
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
Small
Truck
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
Large
Truck
2.5
1-2
0.5
4.5-6.5
9.5-14.5
0.5
Table 1-8 Hybrid Technology Effectiveness
Technology
Stop-Start with 42 volt system
IMA/IS A/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 CO2 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 CO2 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
1-10
-------
Technology Packages, Cost and Effectiveness
1.3 Package Cost and Effectiveness
1.3.1 Explanation of Technology Packages
Individual technologies can be used by manufactures to achieve incremental
CChreductions. 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 CChreductions 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. No commenter took issue with this concept.
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 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
1-11
-------
Regulatory Impact Analysis
order to modify its fleet such that compliance could be achieved. No commenter took issue
with this concept or the list of packages that were developed.
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 COithan gasoline.
As this rule is based on the reduction of CO2emissions 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.
1-12
-------
Technology Packages, Cost and Effectiveness
Figure 1-1 Scaling classes to Vehicle Type Mapping
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 provides superior cost
effectiveness). Table 1-12 shows the package costs for the packages that were actually used
as inputs to 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 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 in 2012 are, on average, 118% 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 109% of those for 2016 as learning has
occurred between 2012 and 2014. The costs for 2022 are 94% 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
GHG 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). No
commenter took issue with this concept.
1-13
-------
Regulatory Impact Analysis
Table 1-11 Package Descriptions and 2016MY Costs for 19 Vehicle Types (T1-T19), All Packages Considered, Costs in 2007 dollars
Vehicle
Subcompact
Car 14
(Type 1)
Compact Car 14
(Type 2)
Midsize Car/Small
MPV (unibody) 14
(Type 3)
Compact
Car/Small MPV
(unibody) V6
(Type 4)
Technology
Package #
101
102
103
104
105
201
202
203
204
205
206
207
301
302
303
304
305
307
308
309
401
402
403
404
406
407
408
Engine
1.5L4VI4
1.5L4VI4 + CCP
1.2L4VI3 + CCP+ DWL + GDI
l.OL 4V 13 (small) Turbo + DCP + GDI
29KWH (FTP 150 miles @160 WH/mi)
2.4L4VI4
2.0L4VI4 + CCP + GDI
2.0L4VI4 + CCP + GDI
2.0L 4V 14 + CCP + DWL + GDI
1.5L4VI4Turbo + DCP + GDI
1.2L4VI4HEV(IMA) + GDI
1.2L4V 14 Plug-in HEV (IMA) + GDI (50% UF)
2.4L4VI4
2.2L4VI4 + CCP + GDI
2.2L4VI4 + CCP+ DWL + GDI
2.2L4VI4 + CCP+ DWL + GDI
1.6L 4V 14 Turbo + DCP + GDI
1.4L 4V 14 Turbo HEV (IMA) + GDI
1.8L4VI4 HEV (Power Split) + GDI
1.8L4VI4 Plug-in HEV (Power Split) + GDI (50% UF)
3.0L4VV6
2.0L 4V 14 Turbo + DCP + GDI
2.0L4VI4Turbo + DCP + GDI
2.0L 4V 14 Turbo + DCP + GDI
2.4L 14 Turbo Diesel
1.5L 4V 14 Turbo HEV (IMA) + GDI
2.8L 4V V6 HEV (2-mode) + CCP + Deac + GDI
Transmission
AT 4 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
N/A
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 6 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
N/A
N/A
AT 4 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
System
Voltage
12V
12V
42V S-S
42V S-S
EV
12V
12V
12V
42V S-S
42V S-S
HEV
HEV
12V
12V
12V
42V S-S
42V S-S
HEV
HEV
HEV
12V
12V
12V
42V S-S
12V
HEV
HEV
Camshaft changes
(not used for downsized
engines)
$
_Q
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
iFriction Rdxn
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
JFuel system
GDI-I4>I3
GDI-I4>I3
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-V6>I4
GDI-V6>I4
GDI-V6>I4
Diesel
GDI-V6>I4
GDI-V6
0)
s |
1 "is
0 c£
Q <
14 to 13
14 to 13
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
14 to 14
Eggressive shift
arly torque lock
ASL TORQ
ASL TORQ
ASL TORQ
ASL TORQ
ASL TORQ
ASL TORQ
V6 DOHC to 14 ASL TORQ
V6 DOHC to 14
V6 DOHC to 14
Diesel-SCR
V6 DOHC to 14
tlternator &
lectrification
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
0}
s
Ul
Q_
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
2
i}
<
AERO 1
AER01
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AER01
AERO 1
AER01
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
ILow RR tires
Low drag brakes
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
JAxle disconnect
Iweight rdxn
3%
5%
10%
20%
3%
3%
10%
10%
3%
5%
10%
10%
3%
3%
5%
5%
12016 MY Cost
$189
$518
$1,205
$1,770
$28,537
$189
$730
$669
$1,475
$1,841
$3,144
$10,066
$189
$749
$908
$1,538
$1,904
$3,602
$4,211
$9,634
$214
$1,111
$1,051
$1,633
$2,722
$4,188
$6,109
1-14
-------
Technology Packages, Cost and Effectiveness
Table 1-11 Continued
Vehicle
Midsize/Large
CarV6
(Type 5)
Midsize Car/Large
CarVS
(Type 6)
Mid-sized MPV
(unibody)/Small
Truck 14
(Type 7)
Jfi
U
3
> i-
Midsize MP
(unibody)/Smal
V6/V8
(Type 8)
Technology
Package #
501
502
503
504
505
508
601
602
603
604
605
606
608
609
701
702
703
704
705
707
708
709
801
802
803
804
805
806
807
808
811
812
813
Engine
3.3L4VV6
3.0L4VV6 + CCP+GDI
3.0L 4V V6 + CCP + Deac + GDI
3.0L4VV6 + CCP+ Deac + GDI
2.2L 4V 14 Turbo + DCP + GDI
2.5L4VI4 HEV (Power Split) + GDI
4.5L4VV8
4.0L4VV6 + CCP+GDI
4.0L 4V V6 + CCP + Deac + GDI
4.0L 4V V6 + CCP+ Deac + GDI
3.0L 4V V6 Turbo + DCP + GDI
3.0L 4V V6 Turbo + DCP + GDI
3.0L 4V V6 Turbo Diesel
3.0L 4V V6 HEV (2-mode) + CCP + Deac + GDI
2.6L4VI4
2.4L4VI4 + CCP + GDI
2.4L4VI4 + CCP+ DWL + GDI
2.4L 4V 14 + CCP + DWL + GDI
2.0L 4V 14 Turbo + DCP + GDI
1.8L 4V 14 Turbo HEV (IMA) + GDI
1.8L4V 14 Turbo HEV (Power Split) + GDI
1.8L4V 14 Turbo Plug-in HEV (IMA) + GDI (50% UF)
3.7L2VSOHCV6
3.2L 2V SOHC V6 + CCP + GDI
3.2L 2V SOHC V6 + CCP + Deac + GDI
2.8L4VV6 + CCP + GDI
2.8L 4V V6 + CCP + DWL + GDI
2.8L4VV6 + CCP+ Deac + GDI
2.8L 4V V6 + CCP + Deac + GDI
2.4L 4V 14 Turbo + DCP + GDI
2.8L 14 Turbo Diesel
3.0L 4V V6 HEV (IMA) + CCP + Deac + GDI
3.0L 4V V6 HEV (2-mode) + CCP + Deac + GDI
Transmission
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
N/A
dry DCT 6 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
System
Voltage
12V
12V
12V
42V S-S
42V S-S
HEV
12V
12V
42V S-S
42V S-S
42V S-S
42V S-S
12V
HEV
12V
12V
12V
42V S-S
42V S-S
HEV
HEV
HEV
12V
12V
12V
12V
12V
12V
42V S-S
42V S-S
12V
HEV
HEV
Camshaft changes
(not used for downsized
engines)
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
SJ
_Q
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
-a
c£
o
IE
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
JFuel system
GDI-V6
GDI-V6
GDI-V6
GDI-V6>I4
GDI-V6>I4
GDI-V8>V6
GDI-V8>V6
GDI-V8>V6
GDI-V8>V6
GDI-V8>V6
Diesel
GDI-V8>V6
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-I4
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6>I4
Diesel
GDI-V6
GDI-V6
8
! ;
V6 DOHC to 14
V6 DOHC to 14
•Aggressive shift
ASL
ASL
ASL
ASL
V8 DOHC to V6 DOHC ASL
V8 DOHC to V6 DOHC ASL
V8 DOHC to V6 DOHC
V8 DOHC to V6 DOHC ASL
V8 DOHC to V6 DOHC
Diesel-SCR
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
ASL
ASL
ASL
ASL
ASL
ASL
ASL
ASL
V6 SOHC to 14
Diesel-SCR
lEarly torque lock
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
tlternator &
lectrification
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
Ipower steering
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
2
i}
<
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AER01
AERO 1
ILow RR tires
Low drag brakes
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
JAxle disconnect
Iweight rdxn
3%
5%
10%
10%
3%
5%
5%
5%
5%
5%
3%
3%
10%
10%
3%
3%
3%
5%
5%
10%
10%
5%
K016 MY Cost
$214
$985
$1,238
$1,919
$1,903
$5,329
$214
$817
$1,567
$1,506
$2,274
$2,214
$3,258
$5,939
$214
$838
$959
$1,788
$2,141
$4,401
$6,104
$13,999
$214
$1,006
$1,156
$1,264
$1,563
$1,531
$2,248
$2,149
$2,770
$4,430
$5,858
1-15
-------
Regulatory Impact Analysis
Table 1-11 Continued
Vehicle
Large MPV (unibody)
V6
(Type 9)
Large MPV (unibody)
V8
(Type 10)
' Large Truck (+ Van) V6
' (Type 11)
Large Truck H
Large MPV Vf
(T12)
Technology
Package #
901
902
903
904
905
906
907
910
911
1001
1002
1003
1004
1005
1006
1007
1010
1011
1101
1102
1103
1104
1105
1106
1107
1108
1111
1112
1113
1201
1202
1203
1204
1205
Engine
4.0L2VSOHCV6
3.6L2VSOHCV6 + CCP + GDI
3.6L 2V SOHC V6 + CCP + Deac + GDI
3.2L4VV6 + CCP + GDI
3.2L4VV6 + CCP+ Deac + GDI
3.2L4VV6 + CCP+ Deac + GDI
2.4L 4V 14 Turbo + DCP + GDI
2.0L4VI4TurboHEV(IMA) + GDI
3.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.7L2VSOHCV8
4.4L2VSOHCV8 + CCP + GDI
4.4L 2V SOHC V8 + CCP + Deac + GDI
4.2L4VV6 + CCP + GDI
4.2L 4V V6 + CCP + Deac + GDI
4.2L 4V V6 + CCP + Deac + GDI
2.8L 4V V6 Turbo + DCP + GDI
3.0LV6 Turbo Diesel
4.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.2L2VSOHCV6
3.9L 2V SOHC V6 + CCP + GDI
3.9L 2V SOHC V6 + CCP + Deac + GDI
3.6L4VV6 + CCP + GDI
3.6L 4V V6 + CCP + DVVL + GDI
3.6L4VV6 + CCP+ Deac + GDI
3.6L4VV6 + CCP+ Deac + GDI
2.5L4VI4Turbo + DCP + GDI
2.8L 14 Turbo Diesel
3.6L 4V V6 HEV (IMA) + CCP + Deac + GDI
3.6L 4V V6 HEV (2-mode)+ CCP + Deac + GDI
3.8L2VOHVV6
3.2L 4V DOHC V6 + CCP + GDI
3.2L 4V DOHC V6 + CCP + Deac + GDI
3.2L 4V DOHC V6 + CCP + Deac + GDI
2.5L4VI4Turbo + DCP + GDI
Transmission
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
System
Voltage
12V
12V
12V
12V
12V
42V S-S
42V S-S
HEV
HEV
12V
12V
12V
12V
12V
42V S-S
42V S-S
12V
HEV
12V
12V
12V
12V
12V
12V
42V S-S
42V S-S
12v
HEV
HEV
12V
12V
12V
42V S-S
42V S-S
Camshaft changes
(not used for downsized
engines)
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 SOHC to V6 DOHC
V6 OHV to V6 DOHC
V6 OHV to V6 DOHC
V6 OHV to V6 DOHC
8
.a
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
-a
c£
o
."y
it
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
E
S I4 V6 SOHC to 14
GDI-V6>I4 V6 SOHC to 14
GDI-V6
a) "S
£ %
e 8
4-J QJ
QJ CTl
< <
ASL
ASL
ASL
ASL
ASL
ASL
GDI-V8 ASL
GDI-V8 ASL
GDI-V8>V6 V8 SOHC to V6 DOHC ASL
GDI-V8>V6 V8 SOHC to V6 DOHC ASL
GDI-V8>V6 V8 SOHC to V6 DOHC
GDI-V8>V6 V8 SOHC to V6 DOHC
Diesel Diesel-SCR
GDI-V8>V6 V8 SOHC to V6 DOHC
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6>I4 V6 SOHC to 14
Diesel
GDI-V6
GDI-V6
ASL
ASL
ASL
ASL
ASL
ASL
Diesel-SCR
ASL
GDI-V6 ASL
GDI-V6 ASL
GDI-V6
GDI-V6>I4 V6 OHV to 14 DOHC
lEarly torque lock
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
tlternator &
lectrification
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
Ipower steering
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
S
3
AER01
AER01
AERO 1
AER01
AER01
AERO 1
AER01
AER01
AER01
AERO 1
AER01
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AERO 1
AERO 1
AER01
AERO 1
ILow RR tires
Low drag brakes
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 LDB
LRR LDB
LRR LDB
LRR LDB
LRR LDB
LRR
LRR
LRR
LRR LDB
LRR LDB
|Axle disconnect
Iweight rdxn
3%
3%
3%
5%
10%
10%
3%
3%
3%
5%
10%
10%
5%
3%
3%
3%
5%
5%
10%
10%
5%
3%
3%
10%
10%
J2016 MY Cost
$214
$1,027
$1,176
$1,285
$1,565
$2,316
$2,217
$4,512
$6,367
$214
$1,092
$1,242
$1,106
$1,401
$2,187
$2,872
$2,839
$6,167
$239
$948
$1,117
$1,206
$1,590
$1,500
$2,430
$2,312
$3,579
$5,545
$8,257
$239
$1,413
$1,581
$2,637
$2,760
1-16
-------
Technology Packages, Cost and Effectiveness
Table 1-11 Continued
Vehicle
Large Truck (+ Van) VS
(Type 13)
S>3
H •? a
8.3 S
jit
Midsize MPV
body)/Small Truck
V6/V8
(Type 15)
c
3
Large MPV
(unibody) V6
(Type 16)
Technology
Package #
1301
1302
1303
1304
1305
1306
1307
1310
1311
1401
1402
1403
1404
1405
1501
1502
1503
1504
1505
1506
1509
1510
1511
1601
1602
1603
1604
1605
1608
1609
Engine
5.7L2VOHVV8
5.2L2VOHVV8 + CCP + GDI
5.2L 2V OHV V8 + CCP + Deac + GDI
4.6L4VV8 + CCP + GDI
4.6L 4V V8 + CCP + Deac + GDI
4.6L 4V V8 + CCP + Deac + GDI
3.5L 4V V6 Turbo + DCP + GDI
3.5LV6 Turbo Diesel
4.6L 4V V8 HEV (2-mode) + CCP + Deac + GDI
5.4L3VSOHCV8
4.6L 4V DOHC V8 + CCP + GDI
4.6L 4V DOHC V8 + CCP + Deac + GDI
4.6L 4V DOHC V8 + CCP + Deac + GDI
3.5L 4V V6 Turbo + DCP + GDI
3.2L4VV6
2.8L4VV6 + CCP + GDI
2.8L 4V V6 + CCP + DVVL + GDI
2.8L4VV6 + CCP+ Deac + GDI
2.8L 4V V6 + CCP + Deac + GDI
2.4L 4V 14 Turbo + DCP + GDI
2.8L 14 Turbo Diesel
3.0L 4V V6 HEV (IMA) + CCP + Deac + GDI
3.0L 4V V6 HEV (2-mode) + CCP + Deac + GDI
3.5L4VV6
3.2L4VV6 + CCP + GDI
3.2L4VV6 + CCP+ Deac + GDI
3.2L 4V V6 + CCP + Deac + GDI
2.4L 4V 14 Turbo + DCP + GDI
2.0L 4V 14 Turbo HEV (IMA) + GDI
3.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
Transmission
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
System
Voltage
12V
12V
12V
12V
12V
42V S-S
42V S-S
12V
HEV
12V
12V
12V
42V S-S
42V S-S
12V
12V
12V
12V
42V S-S
42V S-S
12V
HEV
HEV
12V
12V
12V
42V S-S
42V S-S
HEV
HEV
Camshaft changes
(not used for downsized
engines)
V8 OHV to V8 DOHC
V8 OHV to V8 DOHC
V8 OHV to V8 DOHC
V8 OHV to V8 DOHC
V8 SOHC 3V to V8 DOHC
V8 SOHC 3V to V8 DOHC
V8 SOHC 3V to V8 DOHC
SJ
_Q
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
X
~§L
|
IE
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
* £ 1
E E S S «> J
-H; P~ o ro
s- 1 1 i £ li
1 1 d s f 1 »
u_ Q < < LU ^^^^i
1 *
1 E I 1 x °
« S E § i |
^ csL -a -a -^ ^
£ o o ~x > o
< _J _J < 5 01
ASL TORQ IACC 12V LRR $239
GDI-V8 ASL TORQ IACC 12V AERO 1 LRR 3% $993
GDI-V8 ASL TORQ IACC 12V AERO 1 LRR 3% $1,162
GDI-V8 ASL TORQ IACC 12V AERO 1 LRR 3% $1,541
GDI-V8 ASL TORQ IACC 12V AERO 1 LRR 5% $1,862
GDI-V8 IACC 42V EPS AERO 1 LRR LDB 10% $2,859
GDI-V8>V6 V8 OHV to V6 DOHC IACC 42V EPS AERO 1 LRR LDB 10% $3,314
Diesel Diesel-SCR AERO 1 LRR LDB 5% $3,646
GDI-V8 AERO 1 LRR LDB $8,552
ASL TORQ IACC 12V LRR $239
GDI-V8 ASL TORQ IACC 12V AERO 1 LRR 3% $1,246
GDI-V8 ASL TORQ IACC 12V AERO 1 LRR 5% $1,566
GDI-V8 IACC 42V EPS AERO 1 LRR LDB 10% $2,564
GDI-V8>V6 V8 SOHC 3V to V6 DOHC IACC 42V EPS AERO 1 LRR LDB 10% $2,952
ASL TORQ IACC 12V LRR $214
GDI-V6 ASL TORQ IACC 12V EPS AERO 1 LRR 3% $1,006
GDI-V6 ASL TORQ IACC 12V EPS AERO 1 LRR 5% $1,305
GDI-V6 ASL TORQ IACC 12V EPS AERO 1 LRR 5% $1,273
GDI-V6 IACC 42V EPS AERO 1 LRR 5% $1,697
GDI-V6>I4 V6 DOHC to 14 IACC 42V EPS AERO 1 LRR 5% $1,681
Diesel Diesel-SCR EPS AERO 1 LRR 5% $2,770
GDI-V6 AERO 1 LRR $4,172
GDI-V6 AERO 1 LRR $5,600
ASL TORQ IACC 12V LRR $214
GDI-V6 ASL TORQ IACC 12V EPS AERO 1 LRR 3% $1,027
GDI-V6 ASL TORQ IACC 12V EPS AERO 1 LRR 5% $1,307
GDI-V6 IACC 42V EPS AERO 1 LRR 10% $2,058
GDI-V6>I4 V6 DOHC to 14 IACC 42V EPS AERO 1 LRR 5% $1,715
GDI-V6>I4 V6 DOHC to 14 AERO 1 LRR $4,188
GDI-V6 AERO 1 LRR $6,109
1-17
-------
Regulatory Impact Analysis
Table 1-11 Continued
Vehicle
Large MPV
(unibody) V8
(Type 17)
Large Truck (+ Van) Ve
(Type 18)
Large Truck (+
Van) V8
(Type 19)
Technology
Package #
1701
1702
1703
1704
1705
1708
1709
1801
1802
1803
1804
1805
1806
1809
1810
1811
1901
1902
1903
1904
1905
1908
1909
Engine
4.6L4VV8
4.2L4VV6 + CCP + GDI
4.2L 4V V6 + CCP + Deac + GDI
4.2L 4V V6 + CCP + Deac + GDI
2.8L 4V V6 Turbo + DCP + GDI
3.0LV6 Turbo Diesel
4.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.0L4VV6
3.6L4VV6 + CCP + GDI
3.6L 4V V6 + CCP + DVVL + GDI
3.6L4VV6 + CCP+ Deac + GDI
3.6L4VV6 + CCP+ Deac + GDI
2.5L4VI4Turbo + DCP + GDI
2.8L 14 Turbo Diesel
3.6L 4V V6 HEV (IMA) + CCP + Deac + GDI
3.6L 4V V6 HEV (2-mode) + CCP + Deac + GDI
5.6L4VV8
4.6L4VV8 + CCP + GDI
4.6L 4V V8 + CCP + Deac + GDI
4.6L 4V V8 + CCP + Deac + GDI
3.5L 4V V6 Turbo + DCP + GDI
3.5LV6 Turbo Diesel
4.6L 4V V8 HEV (2-mode) + CCP + Deac + GDI
Transmission
AT 4 spd
AT 8 spd
AT 8 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
DCT 6 spd
N/A
System
Voltage
12V
12V
12V
42V S-S
42V S-S
12V
HEV
12V
12V
12V
12V
42V S-S
42V S-S
12V
HEV
HEV
12V
12V
12V
42V S-S
42V S-S
12V
HEV
Camshaft changes
(not used for downsized
engines)
8
_Q
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
LUB
-a
c£
o
ut
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
EFR
E
S
1 1
cS S
< <
ASL
GDI-V8>V6 V8 DOHC to V6 DOHC ASL
GDI-V8>V6 V8 DOHC to V6 DOHC ASL
GDI-V8>V6 V8 DOHC to V6 DOHC
GDI-V8>V6 V8 DOHC to V6 DOHC
Diesel Diesel-SCR
GDI-V8>V6 V8 DOHC to V6 DOHC
GDI-V6
GDI-V6
GDI-V6
GDI-V6
GDI-V6>I4 V6 DOHC to 14
Diesel
GDI-V6
GDI-V6
ASL
ASL
ASL
ASL
Diesel-SCR
ASL
GDI-V8 ASL
GDI-V8 ASL
GDI-V8
GDI-V8>V6 V8 DOHC to V6 DOHC
Diesel Diesel-SCR
GDI-V8
lEarly torque lock
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
TORQ
tlternator &
lectrification
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
IACC 12V
IACC 12V
IACC 12V
IACC 42V
IACC 42V
CTl
1
Ul
Q_
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
EPS
§
<
AERO 1
AER01
AERO 1
AERO 1
AER01
AERO 1
AER01
AERO 1
AER01
AER01
AERO 1
AER01
AER01
AERO 1
AERO 1
AER01
AER01
AERO 1
AER01
AER01
ILOW RR tires
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
LRR
low drag brakes
LDB
LDB
LDB
LDB
LDB
LDB
LDB
LDB
LDB
Iweight rdxn
3%
5%
10%
10%
5%
3%
5%
5%
10%
10%
5%
3%
5%
10%
10%
5%
12016 MY Cost
$214
$860
$1,155
$1,941
$2,648
$2,839
$5,921
$239
$948
$1,332
$1,242
$2,172
$2,138
$3,579
$5,287
$8,000
$239
$1,033
$1,353
$2,351
$2,807
$3,646
$8,044
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.
1-18
-------
Technology Packages, Cost and Effectiveness
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
Technology
Package
100
101
102
103
104
200
201
203
204
207
300
301
303
304
309
400
401
403
404
406
408
500
501
502
503
505
508
600
601
602
604
609
Engine
1.5L 4V DOHC 14
1.5L4VI4
1.5L4VI4 + CCP
1.2L 4V13 + CCP + DWL + GDI
l.OL 4V13 (small) Turbo + DCP + GDI
2.4L 4V DOHC 14
2.4L 4V14
2.0L 4V14 + CCP + GDI
2.0L 4V14 + CCP + DWL + GDI
1.2L 4V14 Plug-in HEV (IMA) + GDI (50% UF)
2.4L 4V DOHC 14
2.4L 4V14
2.2L 4V14 + CCP + DWL + GDI
2.2L 4V14 + CCP + DWL + GDI
1.8L 4V14 Plug-in HEV (Power Split) + GDI (50% UF)
3.0L 4V DOHC V6
3.0L 4V V6
2.0L 4V14 Turbo + DCP + GDI
2.0L 4V14 Turbo + DCP + GDI
2.4L 14 Turbo Diesel
2.8L 4V V6 HEV (2-mode) + CCP + Deac + GDI
3.3L 4V DOHC V6
3.3L4VV6
3.0L 4V V6 + CCP +GDI
3.0L 4V V6 + CCP + Deac + GDI
2.2L4VI4Turbo + DCP + GDI
2.5L 4V14 HEV (Power Split) + GDI
4.5L 4V DOHC V8
4.5L4VV8
4.0L 4V V6 + CCP+ GDI
4.0L 4V V6 + CCP+ Deac + GDI
3.0L 4V V6 HEV (2-mode) + CCP + Deac + GDI
Transmission
AT 4 spd
AT 4 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
AT 4 spd
AT 4 spd
DCT 6 spd
dry DCT 6 spd
dry DCT 6 spd
AT 4 spd
AT 4 spd
DCT 6 spd
dry DCT 6 spd
N/A
AT 4 spd
AT 4 spd
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
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
N/A
System
Voltage
12V
12V
12V
42V S-S
42V S-S
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
12V
HEV
12V
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
HEV
.31
176
006
063
249
.31
102
387
311
008
$227
$1,143
$1,967
$2,974
$9,230
$227
$1,071
$1,348
$2,263
$5,831
$222
$1,111
$1,804
$2,887
$7,508
$222
$1,041
$1,310
$2,092
$5,658
$218
$1,081
$1,668
$2,803
$6,127
$218
$1,013
$1,274
$1,947
$5,491
$214
$1,051
$1,633
$2,722
$6,109
$214
$985
$1,238
$1,903
$5,329
$214
$1,051
$1,633
$2,722
$6,109
$214
$985
$1,238
$1,903
$5,329
$214
$1,051
$1,633
$2,722
$6,109
$214
$985
$1,238
$1,903
$5,329
$214
$1,051
$1,633
$2,722
$6,109
$214
$985
$1,238
$1,903
$5,329
$214
$1,051
$1,633
$2,722
$6,109
$214
$985
$1,238
$1,903
$5,329
$214
$1,051
$1,633
$2,722
$6,109
$214
$985
$1,238
$1,903
$5,329
$207
$920
$1,452
$2,492
$5,371
$207
$949
$1,193
$1,713
$4,595
7.6%
23.4%
31.6%
32.9%
36.5%
7.6%
17.9%
20.6%
34.3%
37.5%
$231 $227 $222 $218 $214 $214 $214
$912 $887 $863 $839 $817 $817 $817
$1,863 $1,828 $1,670 $1,538 $1,506 $1,506 $1,506
$9,065 $9,052 $7,330 $5,951 $5,939 $5,939 $5,939
$214 $214 $214 $207 7.6%
$817 $817 $817 $772 17.9%
$1,506 $1,506 $1,506 $1,406 31.9%
$5,939 $5,939 $5,939 $5,190 44.4%
1-19
-------
Regulatory Impact Analysis
Table 1-12 Continued
Technology
Package
700
701
703
704
708
709
800
801
802
803
808
813
900
901
902
903
907
911
1000
1001
1004
1006
1011
1100
1101
1102
1103
1108
1200
1201
1202
1204
1205
Engine
2.6L 4V DOHC 14 (15)
2.6L 4V14
2.4L 4V14 + CCP + DWL + GDI
2.4L 4V14 + CCP + DWL + GDI
1.8L 4V14 Turbo HEV (Power Split) + GDI
1.8L 4V14 Turbo Plug-in HEV (IMA) + GDI (50% UF)
3.7L 2V SOHC V6
3.7L 2V SOHC V6
3.2L 2V SOHC V6 + CCP + GDI
3.2L 2V SOHC V6 + CCP + Deac + GDI
2.4L 4V14 Turbo + DCP + GDI
3.0L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.0L 2V SOHC V6
4.0L 2V SOHC V6
3.6L 2V SOHC V6 + CCP + GDI
3.6L 2V SOHC V6 + CCP + Deac + GDI
2.4L 4V14 Turbo + DCP + GDI
3.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.7L 2V SOHC V8
4.7L 2V SOHC V8
4.2L4VV6 + CCP + GDI
4.2L 4V V6 + CCP + Deac + GDI
4.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.2L 2V SOHC V6
4.2L 2V SOHC V6
3.9L 2V SOHC V6 + CCP + GDI
3.9L 2V SOHC V6 + CCP + Deac + GDI
2.5L4VI4Turbo + DCP + GDI
3.8L 2V OHV V6
3.8L 2V OHV V6
3.2L 4V DOHC V6 + CCP + GDI
3.2L 4V DOHC V6 + CCP + Deac + GDI
2.5L 4V14 Turbo + DCP + GDI
Transmission
AT 4 spd
AT 4 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
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
System
Voltage
12V
12V
12V
42V S-S
HEV
HEV
12V
12V
12V
12V
42V S-S
HEV
12V
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
HEV
12V
12V
12V
12V
42V S-S
12V
12V
12V
42V S-S
42V S-S
2012
$231
$1,072
$2,181
$6,884
$21,528
$231
$1,126
$1,295
$2,588
^8,746
$231
$1,149
$1,318
$2,665
^9,541
$231
$1,239
$2,632
^316
$256
$1,057
$1,248
$2^78
$256
$1,582
$3,145
$3,285
2013
2014
$227 $222
$1,042 $1,014
$2,136 $1,969
$6,680 $6,482
;21,504 $17,349
2015
$218
$227
$1,094
$1,259
$2,532
$8,718
$227
$1,117
$1,281
$2,607
19,513
$222
$1,064
$1,223
$2,353
^7,146
$222
$1,086
$1,245
$2,425
67,782
$1,828
$6,290
$14,020
$218
$1,035
$1,189
$2,200
$5,884
$218
$1,056
$1,210
$2,270
$6,392
$227 $222 $218
$1,204 $1,170 $1,138
$2,574 $2,394 $2,240
$9,294 $7,570 $6,187
$252 $247 $243
$1,029 $1,001 $974
$1,213 $1,180 $1,148
$2,721 $2,528 $2,365
$252 $247 $243
$1,537 $1,495 $1,453
$3,076 $2,873 $2,700
$3,212 $3,005 $2,827
2016
2017
$214 $214
$959 $959
$1,788 $1,788
$6,104 $6,104
$13,999 $13,999
$214
$1,006
$1,156
$2,149
$5,858
$214
$1,027
$1,176
$2,217
.,367
$214
$1,006
$1,156
$2,149
$5^58
$214
$1,027
$1,176
$2,217
$6,367
$214 $214
$1,106 $1,106
$2,187 $2,187
.,167 $6,167
$239 $239
$948 $948
$1,117 $1,117
$2^12 $2,312
$239 $239
$1,413 $1,413
$2,637 $2,637
$2,760 $2,760
2018
$214
$959
$1,788
$6,104
$214
$1,006
$1,156
$2,149
$5,858
$214
$1,027
$1,176
$2,217
$6^67
$214
$1,106
$2,187
$6,167
$239
$948
$1,117
$2^12
$239
$1,413
$2,637
$2,760
2019
$214
$959
$1,788
$6,104
$13,999
$214
$1,006
$1,156
$2,149
$5^58
$214
$1,027
$1,176
$2,217
$6^67
$214
$1,106
$2,187
$6,167
$1,117
$2^12
$239
$1,413
$2,637
$2,760
2020
$214
$959
$1,788
$6,104
$13,999
$214
$1,006
$1,156
$2,149
$5^58
$214
$1,027
$1,176
$2,217
$6^67
$214
$1,106
$2,187
$6,167
$239 $239
$1,117
$2^12
$239
$1,413
$2,637
$2,760
2021
$214
$959
$1,788
$6,104
$13,999
$214
$1,006
$1,156
$2,149
$5,858
$214
$1,027
$1,176
$2,217
$6^67
$214
$1,106
$2,187
$6,167
$239
$948
$1,117
$2^12
$239
$1,413
$2,637
$2,760
2022
$207
$917
$1,679
$5,336
jll.985
$207
$970
$1,114
$1,983
$5,177
$207
$990
$1,134
$2,049
$5,619
$207
$1,066
$2,078
$5,427
$231
$914
$1,077
$2^38
$231
$1,362
$2,509
$2,582
o
B
r\l
8
7.6%
21.4%
34.7%
39.6%
62^%
7.6%
17.8%
19.6%
32.3%
36.3%
^m
7.6%
17.4%
19.4%
32.3%
36.5%
^m
7.6%
18.3%
34.3%
36.5%
^m
7.6%
18.3%
19.9%
35.1%
^m
7.6%
18.9%
34.9%
35.1%
1-20
-------
Technology Packages, Cost and Effectiveness
Table 1-12 Continued
Technology
Package
1300
1301
1302
1303
1306
1307
1400
1401
1402
1404
1405
1500
1501
1502
1505
1511
1600
1601
1602
1605
1609
1700
1701
1702
1704
1709
1800
1801
1802
1806
1900
1901
1902
1904
1905
Engine
5.7L 2V OHV V8
5.7L 2V OHV V8
5.2L 2V OHV V8 + CCP + GDI
5.2L 2V OHV V8 + CCP + Deac + GDI
4.6L 4V V8 + CCP + Deac + GDI
3.5L 4V V6 Turbo + DCP + GDI
5.4L 3V SOHC V8
5.4L 3V SOHC V8
4.6L 4V DOHC V8 + CCP + GDI
4.6L 4V DOHC V8 + CCP + Deac + GDI
3.5L 4V V6 Turbo + DCP + GDI
3.2L 4V DOHC V6
3.2L4VV6
2.8L 4V V6 + CCP + GDI
2.8L 4V V6 + CCP + Deac + GDI
3.0L 4V V6 HEV (2-mode) + CCP + Deac + GDI
3.5L 4V DOHC V6
3.5L4VV6
3.2L4VV6 + CCP + GDI
2.4L 4V14 Turbo + DCP + GDI
3.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.6L 4V DOHC V8
4.6L 4V V8
4.2L4VV6 + CCP + GDI
4.2L 4V V6 + CCP + Deac + GDI
4.2L 4V V6 HEV (2-mode) + CCP + Deac + GDI
4.0L 4V DOHC V6
4.0L 4V V6
3.6L 4V V6 + CCP + GDI
2.5L 4V14 Turbo + DCP + GDI
5.6L 4V DOHC V8
5.6L4VV8
4.6L 4V V8 + CCP + GDI
4.6L 4V V8 + CCP + Deac + GDI
3.5L 4V V6 Turbo + DCP + GDI
Transmission
AT 4 spd
AT 4 spd
AT 6 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 8 spd
DCT 6 spd
N/A
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
AT 4 spd
AT 4 spd
AT 6 spd
DCT 6 spd
DCT 6 spd
System
Voltage
12V
12V
12V
12V
42V S-S
42V S-S
12V
12V
12V
42V S-S
42V S-S
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
HEV
12V
12V
12V
42V S-S
12V
12V
12V
42V S-S
42V S-S
2012
$256
$1,107
$1,298
$3,397
$3,910
$256
$1,393
$3,063
$3^01
$231
$1,126
$2,078
$8,454
$256
$1,057
$2,582
2013
2014
$252 $247
$1,077 $1,048
$1,262 $1,228
$3,320 $3,110
$252
$1,355
$2,996
$3^21
$227
$1,094
$2,037
$8,435
$247
$1,317
$2,796
$222
$1,064
$1,873
{6,872
2015
$243
$1,020
$1,194
$2,929
$3^98
$243
$1,281
$2,624
$3^24
$218
$1,035
$1,734
15,618
2016
$239
$993
$1,162
$2,859
$239
$1,246
$2,564
$2^52
$214
$1,006
$1,697
$5,600
2017
$239
$993
$1,162
$2,859
63,314
$239
$1,246
$2,564
$214
$1,006
$1,697
{5,600
2018
$239
$993
$1,162
$2,859
$239
$1,246
$2,564
$2^52
$214
$1,006
$1,697
$5,600
2019
$239
$993
$1,162
$2,859
$3^14
$239
$1,246
$2,564
$2^52
$214
$1,006
$1,697
15,600
2020
$239
$993
$1,162
$2,859
$3^14
$239
$1,246
$2,564
$2^52
$214
$1,006
$1,697
$5,600
2021
$239
$993
$1,162
$2,859
$239
$1,246
$2,564
$2^52
$214
$1,006
$1,697
$5,600
2022
$231
$957
$1,120
$2,724
$3,096
$231
$1,201
$2,439
$2^65
$207
$970
$1,605
$4,929
$252
$1,029
{2,530
$256 $252
$1,152 $1,121
$2,822 $2,763
$3,338 $3,263
$247
$1,001
$247
$1,091
$2,569
$3,055
$243
$974
$2^85
$243
$1,061
$2,405
$2,875
$239
$948
$2^38
$239
$1,033
$2,351
$2,807
$239
$948
{2,138
$239
$948
$2,138
$239 $239
$1,033 $1,033
$2,351 $2,351
$2,807 $2,807
$239
$948
$2^38
$239
$1,033
$2,351
$2,807
$239
$948
$2^38
$239
$1,033
$2,351
$2,807
$239
$948
{2^38
$239
$1,033
$2,351
$2,807
$231
$914
{1,936
7.6%
18.3%
19.9%
34.9%
35^%
7.6%
18.9%
34.9%
35.1%
^
7.6%
17.8%
31.8%
35.8%
.31
149
099
249
.31
)60
354
037
$227
$1,117
$2,057
$9,230
$227
$934
$2,304
$9,024
$222
$1,086
$1,892
$7,508
$222
$908
$2,132
$7,308
$218
$1,056
$1,753
$6,127
$218
$884
$1,986
$5,933
$214
$1,027
$1,715
$6,109
$214
$860
$1,941
$5,921
$214
$1,027
$1,715
$6,109
$214
$860
$1,941
$5,921
$214
$1,027
$1,715
$6,109
$214
$860
$1,941
$5,921
$214
$1,027
$1,715
$6,109
$214
$860
$1,941
$5,921
$214
$1,027
$1,715
$6,109
$214
$860
$1,941
$5,921
$214
$1,027
$1,715
$6,109
$214
$860
$1,941
$5,921
$207
$990
$1,531
$5,371
$207
$814
$1,825
$5,174
7.6%
17.4%
31.6%
36.5%
7.6%
17.4%
33.7%
36.5%
7.6%
18.3%
34.5%
$231 7.6%
$996 18.3%
$2,233 34.4%
$2,597 34.5%
1-21
-------
Regulatory Impact Analysis
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 2016
119%
117%
109%
102%
100%
100%
100%
100%
100%
100%
94%
A number of the packages shown in Table 1-11 are not shown in Table 1-12
because it was determined that those 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 RIA, 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 information listed in the
first seven columns is taken from Table 1-11 and/or Table 1-12. The values in the eighth
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.
1-22
-------
Technology Packages, Cost and Effectiveness
Table 1-14 Evaluation of Technology Packages for Vehicle Type #6
Technology
Package
601
602
603
605
604
606
608
609
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. OL V6 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
Reduc-
tion
0%
3%
5%
5%
5%
5%
5%
0%
Total CO2
Reduction
7.6%
17.9%
28.2%
28.5%
31.9%
32.1%
32.3%
44.4%
Total
2016
Cost
$214
$817
$1567
$2274
$1506
$2214
$3258
$5939
$/delta
CO2%
$28
$59
$73
$288
$(220)
$4,568
$4,128
$223
Remove Package 605
601
602
603
604
606
608
609
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%
32.3%
44.4%
$214
$817
$1567
$1506
$2214
$3258
$5939
$28
$59
$73
($16)
$4,568
$4,128
$223
Remove Package 603
601
602
604
606
608
609
4.5L DOHC 4-Valve V8
4.0L V6 GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3.0LV6TurboDCP + GDI
3. OLV6 Turbo Diesel
3.0L V6 w/ Deac GDI + CCP HEV
AT 4 spd
AT 6 spd
DCT 6 spd
DCT 6spd
DCT 6 spd
2-mode
12V
12V
42S-S
42S-S
12V
HEV
0%
3%
5%
5%
5%
0%
7.6%
17.9%
31.9%
32.1%
32.3%
44.4%
$214
$817
$1506
$2214
$3258
$5939
$28
$59
$49
$4,568
$4,128
$223
Remove Package 606
601
602
604
608
609
4.5L DOHC 4-Valve V8
4.0L V6 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
DCT 6 spd
DCT 6 spd
2-mode
12V
12V
42S-S
12V
HEV
0%
3%
5%
5%
0%
7.6%
17.9%
31.9%
32.3%
44.4%
$214
$817
$1506
$3258
$5939
$28
$59
$49
$4,295
$223
Remove Package 608
601
602
604
609
4.5L DOHC 4-Valve V8
4.0L V6 GDI + CCP
4.0L V6 w/ Deac GDI + CCP
3.0L V6 w/ Deac GDI + CCP HEV
AT 4 spd
AT 6 spd
DCT 6 spd
2-mode
12V
12V
42S-S
HEV
0%
3%
5%
0%
7.6%
17.9%
31.9%
44.4%
$214
$817
$1506
$5939
$28
$59
$49
$356
1-23
-------
Regulatory Impact Analysis
The eighth, 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
CCbemissions. Here, however, the arithmetic difference in the effectiveness of two
technology packages provides the best comparison across packages, since the base
CChemissions 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 CChemission reduction as a 10%
difference between two packages with 27% and 37% 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.
The order of the packages has been rearranged slightly from that in Table 1-11 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 (package
605) is very high, while that for the engine with cylinder deactivation with a dual clutch
transmission (package 604) is negative. The cause of this is that the cost of package 604
is lower than that for package 605. If package 604 can achieve a 31.9% reduction in
COiemissions at a cost of $1,506, then there is no point in considering a package which
only achieves a 28.5% reduction in CC^emissions for a cost of $2,274. Therefore,
package 605 was removed and the calculations were repeated. (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 after removing package 605.
The second set of packages now shows one obvious anomaly. Again, the ratio of
incremental cost to incremental effectiveness for package 604 is negative. This occurs
because package 604 achieves a higher CO2 reduction for less cost than package 603. As
done above, package 603 was eliminated and the calculations were repeated using the
revised set of packages shown in the third section of Table 1-14.
The third set of packages shows another anomaly. Package 606 is roughly a
factor of 10 higher than any of the prior packages. This occurs because package 606
reduces CO2 emissions by only 0.2% over package 604 for an incremental cost of around
$700. A manufacturer would be better off skipping package 606 and moving straight to
package 608 (the diesel) since package 608 has a more attractive (although not much)
ratio than does package 606. Therefore, package 606 was eliminated and the calculations
were repeated using the revised set of packages shown in the fourth section of Table
1-14.
1-24
-------
Technology Packages, Cost and Effectiveness
The greatest anomaly in the fourth set of ratios is that for the diesel package 608.
It is considerably less attractive than package 609 (the 2-mode hybrid). Therefore,
package 608 is removed and results in the list of packages shown in the fifth and last
section of Table 1-14. If EPA believed that manufacturers would prefer to implement
diesel technology over strong hybridization for some reason, both packages could have
been left in the modeling. However, absent such a reason, the diesel engine package was
removed from vehicle type #6. The revised set of technology packages is shown in the
fourth section of Table 1-14.
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 model1. 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 COi 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;
1-25
-------
Regulatory Impact Analysis
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 1-15 below is an example spreadsheet used by EPA to estimate
the synergistic impacts of a technology package for a standard-size car.
1-26
-------
Technology Packages, Cost and Effectiveness
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
Baseline
New
Indicated
Efficiency
Indicated Energy
Brake Energy
Vehicle
Mass
Inertia
Load
13.0%
0%
13.0%
Mech
Efficiency
38.0% 71.1%
Road Loads
Drag
Aero
Load
4.0%
16%
3.4%
Brake
Efficiency
27.0%
31.5%
Tires
Rolling
Load
Parasitics
Access
Losses
4.0% 1.8%
8% 64%
3.7% 0.8%
Drivetrain
Efficiency
Fuel
Efficiency
77.8% 21.0%
87.2% 27.5%
Gearbox,
T.C.
Trans
Losses
4.2%
33%
3.3%
Road
Loads
100.0%
95.4%
Engine Friction
Friction Pumping
Losses Losses
Heat
Lost To
Exhaust &
Coolant
Ind Eff
Losses
6.6% 4.4% 32.0%
16% 75%
5.6% 1.1% 31.8%
Second
Law
30.0%
•
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)
PMEP Brake
Losses Efficiency
11% | 27%
=71.1% mech efficiency
Technology
Aero Drag Reduction
Rolling Resistance Reduction
Low Fric Lubes
EF Reduction
ICP
DCP
CCP
Deac
DVVL
CWL
Camless
GDI
Turbo/Dnsize
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% total VVT
3.0% total VVT
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% over gas
30.0% over gas
8.5%
Loss Category
Aero
Rolling
Friction
Friction
Pumping
Pumping
Pumping
Pumping, friction
Pumping
Pumping
Pumping
Ind Eff
Pumping
Trans, pumping
Trans, pumping
Pumping
Trans
Trans, pumping
Trans
F, P, A
Access
Access
Access
nd. Eff, pumping
nd. Eff, pumping
nd Eff, pumping
nd Eff, pumping
nd. Eff, pumping
User Picklist
Implementation into estimator Include? (0/1) Gross FC Red
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% Ind Eff
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
1 8% access
1 8% access
36% access
41% IE, 25% pumping
40% IE, 38% pumping
48% IE, 85% pumping, -13% friction
46% IE, 80% pumping, -13% friction
39% IE, 40% pumping
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
3.0%
1.5%
0.5%
2.0%
0.0%
0.0% Pick one
3.0%
0.0%
4.0% ,
0.0% Plckone
0.0%
0.0%
0.0%
0.0% Pick one or
0.0% 6-spd
1.5%
0.5%
5.5% \ Or #44/45
6.5%
7.5%
0.0% | Or #53
1.5%
3.0% Or #51
0.0%
0.0%
0.0% Pick one
0.0%
0.0%
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 COi reduction (i.e. disregarding synergies) to the lumped parameter
results. The difference is the implied synergistic effects of these technologies combined
on a package.
1-27
-------
Regulatory Impact Analysis
Table 1-16 Comparison of Lumped Parameter Analysis with Standard Car Package
TECHNOLOGY
Aero Drag
Rolling Resistance Reduction
Low Friction Lubricants
Engine Friction Reduction
VVT - Coupled Cam Phasing
VVT - Discrete Variable Lift
Aggressive Shift Logic
Early Torque Convertor 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
CO2REDUCTION
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. No commenter took issue with this concept.
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.
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
1-28
-------
Technology Packages, Cost and Effectiveness
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 COi 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
COi emissions over the EPA combined cycle federal test procedure (FTP) for five
baseline vehicles, for validation with EPA certification data.
1-29
-------
Regulatory Impact Analysis
• Simulation of the vehicle technology combinations (applied to the baseline
vehicles)
• 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. No
commenter took issue that the Ricardo work was a legitimate way to validate the lumped
parameter methodology, and that it did in fact confirm that methodology's reasonable use
in this rule
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.
1-30
-------
Technology Packages, Cost and Effectiveness
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,
GMC 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.
• 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 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.
1-31
-------
Regulatory Impact Analysis
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.
Table 1-19 Description of Baseline Vehicles
VEHICLE CLASS
Baseline Vehicle
COiEmissions*
(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
dbs)
ETW (Ibs)
GVWR (Ibs)
GCWR (Ibs)
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
Chrysler 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 COi equivalent, taken from EPA adjusted combined fuel economy ratings.
1-32
-------
Technology Packages, Cost and Effectiveness
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. 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.
1-33
-------
Regulatory Impact Analysis
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%)
1-34
-------
Technology Packages, Cost and Effectiveness
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.
1-35
-------
Regulatory Impact Analysis
Table 1-21 Description of the Vehicle Technology Packages Modeled by Ricardo
VEHICLE
CLASS
Standard
Car
a,
§
13
m
L.
n
U
o>
N
C»
3
u.
0,
§
&
n
J
^
u
L.
H
o>
W)
3
TECHNOLOGY
PACKAGE
Baseline
Z
1
2
Baseline
Z
1
2
15
15a
15b
5
Baseline
4
5
Yl
Y2
6a
16
Baseline
4
6b
16
Baseline
9
10
11
12
17
XI
X2
ENGINE
2.4 Liter 14
2.4L 14, PFI
2.4L 14, GDI
2.4L 14, GDI
2.4 Liter 14
2.4L 14, PFI
2.4L 14, GDI
2.4L 14, GDI
1.5LI4, GDI,
Turbo
2.4L 14, GDI
2.4L 14, GDI,
HCCI
1.9L 14, Diesel
3.5 Liter V6
2.2L 14, GDI,
Turbo
2.8L 14, Diesel
3.5L V6, GDI
3.5L V6, GDI,
HCCI
3.0L V6, GDI
3.5L V6, GDI
3.8 Liter V6
2. 1L 14, GDI,
Turbo
3.0L V6, GDI
3.8L V6, GDI
5.4 Liter, V8
5.4L V8, GDI
3.6L V6, GDI,
Turbo
4.8L V8, Diesel
5.4L V8, GDI
5.4L V8, GDI
5.4L V8, GDI
5.4L V8, GDI,
HCCI
VALVETRAIN
DOHC, DCP
CCP, DVVL
DCP, DVVL
DCP
DOHC, DCP
CCP, DVVL
DCP, DVVL
DCP
DCP
CVA
DCP, CVVL
DOHC
SOHC
DCP
DOHC
CVA
DCP, CVVL
DCP, CVVL
CCP, Deac
OHV
DCP
CCP, Deac
CCP, Deac
SOHC, CCP
CCP, Deac
DCP
DOHC
CCP, Deac
DCP, DVVL
CVA
DCP, CVVL
TRANSMISSION
L5
DCT6
CVT
L6
L6
DCT6
CVT
L6
DCT6
DCT6
DCT6
DCT6
L5
L6
DCT6
DCT6
DCT6
DCT6
L6
L4
L6
DCT6
L6
L4
DCT6
DCT6
DCT6
L6
L6
DCT6
DCT6
ACCESSORIES
—
ISO (42V), EPA,
EACC
EPS, EACC, HEA
ISO (42V), EPS,
EACC
EPS
ISO (42V), EPA,
EACC
EPS, EACC, HEA
ISO (42V), EPA,
EACC
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
—
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
ISO (42V), EPA,
EACC
—
EPS, EACC, HEA
EPS, EACC, HEA
ISO (42V), EPA,
EACC
--
ISO (42V), EPA,
EACC
EPS, EACC, HEA
EPS, EACC, HEA
ISO (42V), EPA,
EACC
EPS, EACC, HEA
EPS, EACC, HEA
EPS, EACC, HEA
1-36
-------
Technology Packages, Cost and Effectiveness
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 COi emissions results from the vehicle simulation are summarized below in
Table 1-22 (for cars) and Table 1-23 (for light-duty trucks). The COi 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 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
L.
n
U
0)
N
C»
3
U.
TECHNONOLGY
PACKAGE
Baseline
Z
1
2
Baseline
4
5
Yl
Y2
6a
16
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
GDI, HCCI, DCT
GDI, DCP,
CVVL, DCT
GDI, CCP, Deac,
L6, ISO
C02
CITY
g/mi
338
250
294
277
420
346
315
278
290
331
301
C02
HWY
g/mi
217
170
198
180
279
236
221
199
197
235
205
C02
COMB
g/mi
284
214
251
233
356
296
273
242
248
288
257
C02
REDUCTION
%
--
24.7%
11.5%
17.8%
--
16.9%
23.5%
32.0%
30.4%
19.2%
27.7%
"-Please refer to Table 1-20 for a full description of the vehicle technologies
1-37
-------
Regulatory Impact Analysis
Table 1-23 CO2 Emissions Estimates Obtained from Vehicle Simulation (Light-Duty Trucks)
VEHICLE
OH
S
cd
a
c/3
OH
S
B
VH
cd
hJ
^
O
E§
£P
cd
hJ
TECHNONOLGY
PACKAGE
Baseline
Z
1
2
15
15a
15b
5
Baseline
4
6b
16
Baseline
9
10
11
12
17
XI
X2
MAJOR
FEATURES*
2.4L 14, DCP, EPS
CCP, DVVL, DCT,
ISO
GDI, DCP, DVVL,
CVT
GDI, DCP, L6, ISO
1.5L 14 GDI, Turbo,
DCP, DCT
GDI, CVA, DCT
GDI, HCCI, DCT
1.9L 14 Diesel, DCT
3.8LV6
2.1LI4, GDI, Turbo,
DCP, L6
GDI, CCP, Deac, DCT
GDI, CCP, Deac, L6,
ISO
5.4L V8, CCP
GDI, CCP, Deac, DCT,
ISO
3.6L V6, GDI, Turbo,
DCP, DCT
4.8L V8 Diesel, DCT
GDI, CCP, Deac, L6,
ISO
GDI, DCP, DVVL, L6
GDI, CVA, DCT
GDI, HCCI, DCT
CO2
CITY
g/mi
367
272
310
291
272
262
270
282
458
357
333
325
612
432
404
444
459
492
422
425
CO2
HWY
g/mi
253
208
227
211
212
193
197
205
313
256
248
225
402
315
319
326
328
333
314
311
CO2
COMB
g/mi
316
243
272
255
245
231
237
247
393
312
295
280
517
379
366
391
400
420
374
374
CO2
REDUCTION
%
—
23.0%
13.7%
19.3%
22.5%
26.8%
24.8%
21.8%
—
20.6%
24.9%
28.7%
—
26.7%
29.3%
24.4%
22.6%
18.8%
27.8%
27.7%
"-Please refer to Table 1-20 for a full description of the vehicle technologies
1.5 Comparison of Lumped-Parameter Results to Modeling Results
Considering the following:
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.
1-38
-------
Technology Packages, Cost and Effectiveness
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. No commenter took issue with this concept.
35% -
c
_o
u
CJ
fa
15% -
10% -
5% --
rl f
•
_
333
+3+3+3
Z 1 2
-
r
N Sin MPV
~ Sin MPV
-i
> >
in OH
S 2
~ B
o w
2 15
-,
-, -,
r
^ ^ ^ OOOOOOOOOOOOgj
OTOTOTJJJJJJ §"
15a 15b 5 4 5 6a 16 Yl Y2 4
-
-i
-i
-i i-t
^^|||||||
«D «D
OO OO
6b 16 9 10 11 12 17 XI X2
Package
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
1-39
-------
Regulatory Impact Analysis
some technologies by comparing and isolating Ricardo package results across different
vehicle platforms.
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.
1-40
-------
Technology Packages, Cost and Effectiveness
Engine Technology
VVT(ICP)
I VVT(CCP)
DISP
IVVLT
i
DVVL)|
GDI
TURB
Trans Technology
(Lines indicate potential synergies)
Figure 1-3 Illustration of technology pairings for a specific technology application path
1-41
-------
Regulatory Impact Analysis
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.
3 NHTSA 2008-2011 CAFE FRM at 71 FR 17566. Docket Number: EPA-HQ-OAR-
2009-0472-0162
1-42
-------
Air Conditioning
CHAPTER 2: Air Conditioning
2.1 Overview of Air Conditioning Impacts and Technologies
Over 95% of the new cars and light tracks in the United States are equipped with
mobile air conditioning (A/C) systems. In the 1970's and 1980's, A/C systems were an
optional (luxury) feature, but these systems are now standard on almost all new vehicle
models. The 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 (GHG) directly into the air - even when the vehicle is not in operation.
No other vehicle system has associated GHG leakage. 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 (HFC) refrigerant compound currently
used in all recent model year vehicles is R134a (also known as 1,1,1,2-Tetrafluoroethane,
or HFC-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
A The global wanning 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.
2-1
-------
Regulatory Impact Analysis
1,430 grams of CO2 (which has a GWP 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/designs 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 CO2 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 CC^ 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
2-2
-------
Air Conditioning
discusses some similar features and others that emphasize additional emission reduction
mechanisms.
2.2 Air Conditioner Leakage
No substantive public comments were received during the public comment period on the
size of the HFC credit or the HFC inventories presented here. Consequently, the NPRM
analysis is presented here unchanged. The NPRM inventories differ slightly from the
updated emission inventory analysis provided in the FRM (RIA Chapter 5) due to slight
changes in sales and VMT. The global warming potentials used in this analysis are
discussed in RIA chapter 5.
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
2-3
-------
Regulatory Impact Analysis
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 CO2,
NiO and CH4 emissions are from MOVES, the refrigerant emissions are from the
Vintaging model, and the A/C COi emissions are from EPA and the National Renewable
Energy Laboratory (NREL) as described below.
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
Multiplied by 1430 GWP to Convert to CO2 Equivalent
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
B EPA reported the MVAC emissions at 56.6 Tg COi EQ, using a GWP of 1300. This number has been
adjusted using a GWP of 1430.
2-4
-------
Air Conditioning
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 Approach to Quantify Leakage?
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 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. In addition, at time of
certification, manufacturers are required to attest that the components used in these
systems are durable. 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.
As discussed in the preamble and Response to Comments document, 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 will rely on the best
available design metrics to quantify system performance. A few commenters suggested
that we allow manufacturers, as an option, to use an industry-developed "mini-shed" test
procedure (SAE J2763 - Test Procedure for Determining Refrigerant Emissions from
Mobile Air Conditioning Systems) to measure and report annual refrigerant leakage.0
However, while EPA generally prefers performance testing, for an individual vehicle A/C
c
Honeywell and Volvo supported this view; most other commenters did not.
2-5
-------
Regulatory Impact Analysis
system or component, there is not a strong inherent correlation between a performance
test using SAE J2763 and the design-based approach we are adopting (based on SAE
J2727, as discussed below).0 Establishing such a correlation would require testing of a
fairly broad range of current-technology systems in order to establish the effects of such
factors as production variability and assembly practices (which are included in J2727
scores, but not in J2763 measurements). To EPA's knowledge, such a correlation study
has not been done. At the same time, as discussed below, there are indications that much
of the industry will eventually be moving toward alternative refrigerants with very low
GWPs. EPA believes such a transition would diminish the value of any correlation
studies that might be done to confirm the appropriateness of the SAE J2763 procedure as
an option in this rule. For these reasons, EPA is therefore not adopting such an optional
direct measurement approach to addressing refrigerant leakage at this time. 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
rule (expressed 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:
Equation 1 - Credit Equation
A/C Credit = (MaxCredit) * [ 1 - (§86.166-12 Score/AvgImpactE) *
(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
D However, there is a correlation in the fleet between J2763 measurements and J2727 scores.
E Section 86.166-12 sets out the individual component leakage values based on the SAE value.
2-6
-------
Air Conditioning
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 COi EQ)
A/C credit maximum (grams /mile CO2 EQ)a
§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
Lowest-GWP
Refrigerant
(GWP=1)
Cars
13.8
13.8
8.3
16.6
Trucks
17.2
17.2
10.4
20.7
IWith electric compressor, value increases to 9.5 and 11.7 for cars and trucks, respectively.
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 (VMT).F 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 RIA 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.4'7 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),
F 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.
2-7
-------
Regulatory Impact Analysis
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 §86.112), as a design based test, decreases as the amount of expected
leakage diminishes. 50% of the base rate is therefore 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 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.
2-8
-------
Air Conditioning
AIAM commented that EPA should not set a lower limit on the leakage score,
even for non-electric compressors. EPA has determined not to do so. First, although
there do exist vehicles in the Minnesota data with lower scores than our proposed (and
now final) minimum scores, there are very few car models that have scores less than 8.3,
and these range from 7.0 to about 8.0 and the difference are small compared to our
minimum score.0 More important, lowering the leakage limit would necessarily increase
credit opportunities for equipment design changes, and EPA believes that these changes
could discourage the environmentally optimal result of using low GWP refrigerants.
Introduction of low GWP refrigerants could be discouraged because it may be less costly
to reduce leakage than to replace many of the A/C system components. Moreover, due to
the likelihood of in-use factors, even a leak-less (according to J2727) R134a system will
have some emissions due to manufacturing variability, accidents, deterioration,
maintenance, and end of life emissions, a further reason to cap the amount of credits
available through equipment design. The only way to guarantee a near zero emission
system in-use is to use a low GWP refrigerant. The EPA has therefore decided for the
purposes of this final rule to not change the minimum score for belt driven compressors
due to the reason cited above and to the otherwise overwhelming support for the program
as proposed from commenters.
In addition, as discussed above, EPA recognizes that substituting a refrigerant
with a significantly lower GWP will be a very effective way to reduce the impact of all
forms of refrigerant emissions, including maintenance, accidents, and vehicle scrappage.
To address future GHG regulations in Europe and California, systems using alternative
refrigerants - including HFO1234yf, with a GWP of 4 and CO2 with a GWP of 1 - are
under serious development and have been demonstrated in prototypes by A/C component
suppliers. The European Union has enacted regulations phasing in alternative refrigerants
with GWP less than 150 starting this year, and the State of California proposed providing
credits for alternative refrigerant use in its GHG rule. Within the timeframe of MYs
2012-2016, EPA is not expecting widespread use of low-GWP refrigerants. However,
EPA believes that these developments are promising, and, as proposed, has included in
the A/C Leakage Credit formula above a factor to account for the effective GHG
reductions that could be expected from refrigerant substitution. The A/C Leakage Credits
that will be available will be a function of the GWP of the alternative refrigerant, with the
largest credits being available for refrigerants with GWPs at or approaching a value of 1.
For a hypothetical alternative refrigerant with a GWP of 1 (e.g., COi as a refrigerant),
effectively eliminating leakage as a GHG concern, our credit calculation method could
result in maximum credits equal to total average emissions, or credits of 13.8 and 17.2
g/mi CCheq for cars and trucks, respectively, as incorporated into the A/C Leakage Credit
formula above as the "MaxCredit" term.
0 The Minnesota refrigerant leakage data can be found at
http://www.pca.state.mn.us/cliniatechange/mobileair.htmltleakdata
2-9
-------
Regulatory Impact Analysis
A final adjustment was made to each credit to account for the difference between
real-world HFC emissions and test-cycle COi 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 COi emissions.8 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 COi 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
C02 EQ)
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 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 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.9 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, 2003.10 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
2-10
-------
Air Conditioning
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 ale 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 Minimum 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
penetrations within the 2012-2016 timeframe, as this technology is likely to accompany
hybrid vehicle and stop-start technologies, and not conventional vehicles. However,
several commenters noted that electric A/C compressors are an enabler to lower leakage
rates - beyond the minimum levels we specified - and when this technology is used in
conjunction with other leakage-reducing technologies, the resulting system leakage can
be lower than the minimum levels we proposed (8.3 g/yr for cars and 10.4 g/yr for
trucks). We agree with the commenters that it is feasible for A/C systems with electric
compressors to achieve lower leak rates than belt-driven compressors. Since compressor
leakage can be responsible for more than 50% of the refrigerant leakage from a system,
we are lowering the minimum leakage score for cars and trucks with electric compressors
by 50%, to 4.1 and 5.2 g/yr respectively. The effect of this change will be that vehicles
2-11
-------
Regulatory Impact Analysis
with electric compressors will be able to qualify for credit, on a grams-per-mile basis,
than would have been possible with the limitations of the original minimum leakage
score. For vehicles which do use an electric compressor, the 8.3 and 10.4 g/yr minimum
leakage scores for cars and trucks are retained.
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.40 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 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.11 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
2-12
-------
Air Conditioning
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 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.12 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.13 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
2-13
-------
Regulatory Impact Analysis
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-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.
2-14
-------
Air Conditioning
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
2.2.3.4 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.14 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.
2-15
-------
Regulatory Impact Analysis
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 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 leakage that are modeled: the "normal" leakage that
results in annual refrigerant loss, and the "avoidable" leakage which results in total
refrigerant loss due to failure of the A/C components (e.g. evaporator, condenser, or
compressor). This model is developed to help us estimate the costs of the A/C leakage
reductions. It is especially needed to determine the period over which the discounted cost
savings should be applied.11
Normal refrigerant leakage 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 leakage rates
decrease (proportionally) as the base leakage rates are decreased with the use of
improved parts and components. The base leakage rate is modeled as a linear function,
TT
Air conditioning leakage controls are the only technology in this rule that have 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.
2-16
-------
Air Conditioning
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 in vintaging, 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.
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.
2-17
-------
Regulatory Impact Analysis
.>
O)
_o
'35
120
100
80
•| 60
LU
40
0)
CT
OS
n
0)
-*— leakage
-a— remaining charge
recharge
required
800
700 «•
E
600 S
3
4- 500 CT
c
400-§
300 £
o>
4- 200 E
n
100"
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 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.
1 9O
I £\}
1 00
I UU
'E 80
>> ou
s
o> fin
O) DU
re
^
m ACt
w tu
90
n
•
• •
» •
•
A A
• 1 8 g/yr
• 13.5 g/yr
A Q g/yr
• 4.5 g/yr
A
* .
A A . • '
• A •
A • « "
: . • •
0 2 4 6 8 10
Age of Vehicle
12 14
Figure 2-2 A/C Refrigerant Leakage Rate for Different Technologies as Vehicles Age
2-18
-------
Air Conditioning
pnn
^-v OUU
c vnn
re 700
^ ROD
0)
c
're
E/inn
tuu
a:
c
2 .»,»,»
200
O)
'c 1 nn
o>
cc o
1 i I • . .
I ^ A ' .
A •
» • A
» 18 g/yr
• 13.5 g/yr
A 9 g/yr
• 4.5 g/yr
•
• A
refill line «
A
\j n i i i i
0 2 4 6 8 10
Age (years)
I
12 14
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.15 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.16 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).16
2-19
-------
Regulatory Impact Analysis
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.
2.3 CO2 Emissions due to Air Conditioners
2.3.1 Impact of Air Conditioning Use on Fuel Consumption and
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.17'18 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 comforf'-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.19 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.
2-20
-------
Air Conditioning
The NESCAFF results were taken from Table 3-1 of their report and are
r\r\
summarized in Table 2-7.
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%
NESCAFF 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 CCh 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
COi emissions and fuel use are equivalent, since the type of fuel being used is always
gasoline.1
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.20 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.J 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.
1 Because NESCCAF estimated A/C fuel use nationwide, while ARE focused on that in California, the
NESCCAF and EPA methodologies and results are coempared below.
J 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.
2-21
-------
Regulatory Impact Analysis
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 Track
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 ralemaking which revised the onroad fuel economy labeling procedures
for new motor vehicles.21 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 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
2-22
-------
Air Conditioning
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.8 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.22 Resulting in 14.3 g/mi
per vehicle COi -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).K
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 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
K 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.)
2-23
-------
Regulatory Impact Analysis
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 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
2-24
-------
Air Conditioning
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.
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
2-25
-------
Regulatory Impact Analysis
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 SC03.23 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 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
2-26
-------
Air Conditioning
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).8
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 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.
2-27
-------
Regulatory Impact Analysis
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 CO2 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. 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.
2-28
-------
Air Conditioning
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
CC>2 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 COi per mile. These figures were taken from Tables B-20 through B-23 of the
NESCCAF report.24 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 CC>2 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 System on Fuel Use
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 CCh 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 of A/C on onroad fuel consumption, the
2-29
-------
Regulatory Impact Analysis
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 be 2.6-to-4.1%.
2.3.2 Technologies That Improve Efficiency of Air Conditioning and Their
Effectiveness
EPA estimates that the COi 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 CO2 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 finalizing a program that will 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 CO2 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.
2-30
-------
Air Conditioning
These baseline systems are also extraordinarily wasteful in their energy consumption
because they add heat to the cooled air 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 finalizing will encourage the
reduction of A/C CC^ emissions from cars and trucks by up to 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.25 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.
2-31
-------
Regulatory Impact Analysis
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 2.3.2.1. 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 its supply air from the cooler air inside the vehicle (also
known as recirculated air, or '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 of 35-to-40% improvement is
possible under the conditions of this test. 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
specifying 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 control the air
supply in a 'closed-loop' manner, equipping their A/C systems with humidity sensors or
fog sensors (which detect condensation on the inside glass), allowing them to adjust the
blend of fresh-to-recirculated air and optimize the controls for maximum efficiency. In
response to comments concerning the allowance of additional credit for humidity sensors
(i.e. closed-loop control), we are redefining the credit available for recirculated cabin air
based on how the air supply is controlled. Vehicles with closed-loop control of the air
supply (i.e. sensor feedback is used to control the interior air quality) will qualify for a
1.7 g/mi COi credit and vehicles with open-loop control (sensor feeback is not used to
control interior air quality) will qualify for a 1.1 g/mi COi credit. We believe that the
closed-loop control system will be inherently more efficient than the open-loop control
system because the former can maximize the amount to recirculation to achieve a desired
air quality, whereas the latter will use a fixed 'default' amount of recirculated air which
provides the desired air quality under worst case conditions (e.g. maximum number of
passengers in the vehicle).
2-32
-------
Air Conditioning
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) controller on the 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 either the
blower or fan motors, an overall improvement in A/C system efficiency of 15% is
possible.26 We changed the definition for this credit from requiring waste heat reducing
control on both the blower and fan motors to requiring it 'only' on the blower motor -
whether or not similar control is used on the fan motor. This change was made because
commenters noted the majority of the efficiency gain due to waste-heat-reducing control
technology is realized on the blower motor, and not the fan motor. Since the blower
motor is consuming energy almost 100% of the time (whether for heating, cooling, or
ventilation, the motor is usually running at some speed whenever the vehicle is being
driven), the efficiency to be gained from improved control technology is greatest on this
motor, and thus credit for waste heat reducing technology will apply only when it is used
on the blower motor.
2.3.2.5 Internal Heat Exchanger
An internal heat exchanger (IHX), which is alternatively described as a suction line heat
exchanger, transfers heat from the high pressure liquid entering the evaporator to the gas
exiting the evaporator, which reduces compressor power consumption and improves the
efficiency of the A/C system. Previously, we considered that IHX technology would be
required with the changeover to an alternative refrigerant such as HFO-1234yf, as the
different expansion characteristics of that refrigerant (compared to R-134a) would
necessitate an IHX. However, several commenters noted that an IHX can be used on R-
134a systems as well, and that a significant efficiency improvement can be realized in
doing so. It is estimated that use of an IHX can improve the coefficient of performance
(COP) for the system can be improved by 7%, resulting in a fuel consumption reduction
of l-to-2%.27 EPA believes that a 20% improvement in efficiency relative to the baseline
configuration can be realized if the system includes an IHX, and a 1.1 g/mi credit for an
IHX will be added to the list of efficiency improving technologies.
2.3.2.6 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 by an internal temperature reference to assure a constant temperature level for
2-33
-------
Regulatory Impact Analysis
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 estimated that when fully
developed, calibrated, and optimized to the requirements of particular system design, use
of EXV technology could result in a 20% efficiency improvement over the baseline TXV
system. However, many commenters stated that the EPA estimate for EXV efficiency
was over-stated, that no manufacturers were developing this technology within the
timeframe of this rulemaking, and that it should not be included on the list of efficiency-
improving technologies. These commenters noted that the SAE IMAC report (from
which we referenced the expected efficiency improvement) utilized an EXV in
conjunction with a more efficient compressor - and not as a standalone technology.
Given the uncertainty in the effectiveness of EXV technology, and the statements that no
manufacturers plan on utilizing it, we are removing this technology from the list of
efficiency improving technologies and credits.
2.3.2.7 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.28 We
solicited comments as to how we should define the "baseline" evaporator and condenser
designs which are compared to the "improved" design. The bench test based engineering
analysis that a manufacturer will submit at time of certification. We will consider the
baseline component to be the version which a manufacturer most recently had in
production on the same vehicle or a vehicle in a similar EPA vehicle classification. The
design characteristics of the baseline component (e.g. tube
configuration/thickness/spacing and fin density) are to be documented in an engineering
analysis and compared to the improved components, along with data demonstrating the
2-34
-------
Air Conditioning
COP improvement. This same engineering analysis can be applied to evaporators and
condensers on other vehicles and models (even if the overall size of the heat exchanger is
different), as long as the design characteristics of the baseline and improved components
are the same. 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.
2.3.2.8 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
90
of these devices, the overall system efficiency is reduced. 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 SAEIMAC team estimated that
overall system COP could be improved by 8% if an oil separator was used.11 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. Whether the
oil separator is a standalone component or is integral to the compressor design,
manufacturers can submit an engineering analysis to demonstrate the effectiveness of the
oil separation technology.
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 SAE IMAC 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 CO2 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 COi
credit. For model years 2014 and later, the design-based approach will be supplemented
with a vehicle performance test. In particular, EPA is specifying 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
2-35
-------
Regulatory Impact Analysis
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
has modified 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 has revised 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 CC>2 a vehicle
generated due to A/C use on the Idle Test.30 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 average A/C COi 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
2-36
-------
Air Conditioning
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 2-13). 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 CO2 result.
40.0
35.0
30.0
25.0
Average A/C CO2 rate = 21.3 g/min
\ • •
20.0
o
15.0
10.0
5.0
0.0
10
20
30 40
Test Number
50
70
Figure 2-4 EPA A/C Idle Test Results from Various Vehicle Model Types
Part of this variation in the 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
2-37
-------
Regulatory Impact Analysis
of their vehicles which produced a below-average A/C CO2 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 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. Many
commenters questioned the ability of the A/C Idle Test to measure the effect of certain
A/C technologies (e.g. technologies which improve performance under higher cooling
load conditions), and stated that the test was not representative of real-world driving
conditions. While we acknowledge that there are limitations to the Idle Test, we have
determined that it is still a valid tool evaluating the efficiency of a vehicle's A/C system
under conditions encountered in daily driving. Moreover, we believe that a performance
test is necessary to assure that efficiency-improving technologies are implemented
properly and that the vehicle's A/C system operates in an efficient manner under idle
conditions. In the future, EPA will continue to work with industry groups,
manufacturers, component suppliers, and other government organizations to develop a
procedure for determining A/C system efficiency which incorporates the appropriate test-
bench, modeling, and drive cycle tools. The goal of this exercise is the development of a
reliable, accurate, and verifiable assessment and testing method which minimizes a
manufacturers testing burden. This effort could include component-level assessment of
A/C technologies, modeling of system control strategies, and development of a vehicle-
based test procedure for validating the findings of component-level and system modeling
analyses.
The average A/C COi 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 the full credit, it
will be necessary for each model to achieve an A/C CC>2 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.
Several commenters disagreed with the EPA's threshold for full credit, arguing that the
30% improvement was too aggressive. However, EPA test results on three vehicle size
classes (large car, SUV, and pickup truck) indicate that significant reductions in fuel
consumption can be achieved by simply switching A/C control from outside air (OSA) to
2-38
-------
Air Conditioning
recirculated cabin air. As shown in , the percentage reduction in the CO2 due to A/C use
was greater than 30% in all three cases.
Table 2-12 Effect of Outside Air and Recirculated Cabin Air on A/C Idle Test Results (EPA Testing)
Vehicle Type
Large Car
SUV
Pickup Truck
A/C CO2 Re
w/Outside Air
25.9
17.4
14.1
isult (g/min)
w/Recirc Cabin Air
14.0
11.4
9.0
Change in A/C CO2
w/Recirc (%)
-45.9
-34.5
-36.2
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. However, to avoid having an all-or-nothing threshold of 14.9 g/min on
the Idle Test to qualify for credits, EPA will allow amount of credit to be scaled to Idle
Test result, with vehicles achieving 14.9 g/min or better receiving full credit, and vehicles
achieving 21.3 g/min or higher receiving no credit, as shown in Figure 2-5.
2-39
-------
Regulatory Impact Analysis
i
0,9
0,8
0,7
•2 0,6
•*-<
c
cu
I 0,5
T3
2 0.4
T3
(U
^ 0.3
0,2
0,1
14.0 15.0 16.0 17.0 18.0 19.0
A/C Idle Test Result {g/min.J
20.0
21.0
22.0
Figure 2-5 A/C Credit Adjustment Factor.
Once manufacturers begin using the technologies described in Table 2-13 - 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 CO2 '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 2-13. 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 COi -
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
2-40
-------
Air Conditioning
from 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.31 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.
2-41
-------
Regulatory Impact Analysis
Table 2-13 Efficiency-Improving A/C Technologies and Credits
Technology Description
Estimated Reduction
in A/C CO2
Emissions
A/C Credit (g/mi
CO2)
Reduced reheat, with externally-controlled,
variable-displacement compressor
30%
1.7
Reduced reheat, with externally-controlled,
fixed-displacement or pneumatic variable
displacement compressor
20%
1.1
Default to recirculated air with closed-loop
control of the air supply (sensor feedback to
control interior air quality) whenever the
outside ambient temperature is 75 °F or
higher (although deviations from this
temperature are allowed if accompanied by
an engineering analysis)
30%
1.7
Default to recirculated air with open-loop
control of the air supply (no sensor
feedback) whenever the outside ambient
temperature is 75 °F or higher (although
deviations from this temperature are
allowed if accompanied by an engineering
analysis)
20%
1.1
Blower motor control which limit wasted
electrical energy (e.g. pulsewidth modulated
power controller)
15%
0.9
Internal heat exchanger (or suction line heat
exchanger)
20%
1.1
Improved evaporators and condensers (with
engineering analysis on each component
indicating a COP improvement greater than
10%, when compared to previous design)
20%
1.1
Oil Separator (internal or external to
compressor)
10%
0.6
2-42
-------
Air Conditioning
The estimates for the percent reduction in A/C CCh for each technology are based
in part on the results of SAEIMAC 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 CC>2
emissions for 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 CO2 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.L 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 2-14 shows how these costs may be distributed on a year by year basis
as the 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 $147.32 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
L Author unknown, Alternative Refrigerant Assessment Workshop, SAE Automotive Alternative
Refrigerant Symposium, Arizona, 2003.
2-43
-------
Regulatory Impact Analysis
these potential maintenance savings are not included in the cost and benefit analysis
presented in Chapters 6 and 8 of this RIA. 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.
For indirect CC>2 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.M The
IMAC Efficiency Improvement team of the Society of Automotive Engineers realized an
efficiency improvement of 36.4% based on existing technologies and processes.25 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 for the 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 RIA, the indirect AC fuel
savings has been included in the total fuel savings resulting from this rulemaking.
Table 2-14 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.
M 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.
2-44
-------
Air Conditioning
Table 2-14 Estimated Costs in Each Model Year for New AC Technology, 2007 Dollars
Penetration
AC Leakage (Direct)
AC Indirect
Total
2012
28%
$5
$15
$20
2013
40%
$7
$21
$28
2014
60%
$11
$32
$42
2015
80%
$14
$42
$56
2016
85%
$15
$45
$60
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 RIA 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 2-15).
Table 2-15 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
28%
61.1%
38.9%
1.8
1.6
3.4
2.2
1.4
3.8
3.5
2013
40%
61.9%
38.1%
2.5
2.3
4.8
3.1
2.3
5.4
5.0
2014
60%
63.2%
36.8%
3.8
3.4
7.2
4.7
3.4
8.1
7.5
2015
80%
64.6%
35.4%
5.0
4.6
9.6
6.2
4.6
10.8
10.0
2016
85%
65.6%
34.4%
5.4
4.8
10.2
6.6
4.8
11.5
10.6
2-45
-------
Regulatory Impact Analysis
References"
1IPCC, Chapter 2, "Changes in Atmospheric Constituents and in Radiative Forcing,"
September, 2007. This document is available in Docket EPA-HQ-OAR-2009-0472-0117.
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. This document is available in Docket EPA-HQ-OAR-2009-
0472-0157.
3 Vincent, R., Cleary, K., Ayala, A., Corey, R., "Emissions of HFC-134a from Light-
Duty Vehicles in California," SAE 2004-01-2256, 2004. This document is available in
Docket EPA-HQ-OAR-2009-0472-0186.
4 EPA, 2009, "Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2007,"
http://www.epa.gov/climatechange/emissions/usinventorvreport.html. This document is
available in Docket EPA-HQ-OAR-2009-0472.
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. This document is available
in Docket EPA-HQ-OAR-2009-0175.
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. This
document is available in Docket EPA-HQ-OAR-2009-0472-0178.
7 Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy,
"Transportation Energy Data Book: Edition 27," 2008. This document is available in
Docket EPA-HQ-OAR-2009-0472
8 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. This document is available in Docket EPA-HQ-
OAR-2009-0472.
9 Ikegama, T., Kikuchi, K., "Field Test Results and Correlation with SAEJ2727,"
Proceedings of the SAE 7th Alternative Refrigerant Systems Symposium, 2006. This
document is available in Docket EPA-HQ-OAR-2009-0472.
10 Atkinson, W., Baker, J., Ikegami, T., Nickels, P., "Revised SAEJ2727: SAE Interior
Climate Control Standards Committee Presentation to the European Commission," 2006.
This document is available in Docket EPA-HQ-OAR-2009-0472.
2-46
-------
Air Conditioning
11 Society of Automotive Engineers, "IMAC Team 1 - Refrigerant Leakage Reduction,
Final Report to Sponsors," 2006. This document is available in Docket EPA-HQ-OAR-
2009-0472-0188.
12 Society of Automotive Engineers Surface Vehicle Standard J2727, issued August,
2008, http://www.sae.org. This document is available in Docket EPA-HQ-OAR-2009-
0472-0160.
13 Minnesota Pollution Control Agency, "Model Year 2009 Leakage Rate List,"
http://www.pca.state.mn.us/climatechange/mobileair.html. This document is available in
Docket EPA-HQ-OAR-2009-0472-0161.
14 Society of Automotive Engineers, "IMAC Team 4 - Reducing Refrigerant Emissions
at Service and Vehicle End of Life," June, 2007. This document is available in Docket
EPA-HQ-OAR-2009-0472.
15 Schwarz, W., "Emission of Refrigerant R-134a from Mobile Air-Conditioning
Systems," Study conducted for the German Federal Environmental Office, September,
2001. This document is available in Docket EPA-HQ-OAR-2009-0472.
16 "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.
This document is available in Docket EPA-HQ-OAR-2009-0472-0167. This document is
available in Docket EPA-HQ-OAR-2009-0472-0167.
17 Johnson, V., "Fuel Used for Vehicle Air Conditioning: A State-by-State Thermal
Comfort-Based Approach," SAE 2002-01-1957, 2002. This document is available in
Docket EPA-HQ-OAR-2009-0472-0179.
18 Rugh, J., Johnson, V., Andersen, S., "Significant Fuel Savings and Emission
Reductions by Improving Vehicle Air Conditioning," Mobile Air Conditioning Summit,
Washington DC., April 14-15, 2004. This document is available in Docket EPA-HQ-
OAR-2009-0472-0179.
19 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. This
document is available in Docket EPA-HQ-OAR-2009-0472.
20 "Reducing Greenhouse Gas Emissions from Light-Duty Motor Vehicles," Northeast
States Center for a Clean Air Future, September, 2004. This document is available in
Docket EPA-HQ-OAR-2009-0472.
21 Fuel Economy Labeling of Motor Vehicles: Revision to Improve Calculation of Fuel
Economy Estimates; Final Rule, 71 FR 77872, December 27, 2006. This document is
available in Docket EPA-HQ-OAR-2009-0472.
2-47
-------
Regulatory Impact Analysis
22 EPA Technical Support Document for Fuel Economy Label Regulations, November,
2006. This document is available in Docket EPA-HQ-OAR-2009-0472.
23 Nam, E.K., "Understanding and Modeling NOx Emissions from Automobiles During
Hot Operation," PhD Thesis, University of Michigan, 1999. This document is available in
Docket EPA-HQ-OAR-2009-0472.
24 "Reducing Greenhouse Gas Emissions from Light-Duty Motor Vehicles," Northeast
States Center for a Clean Air Future, September, 2004. This document is available in
Docket EPA-HQ-OAR-2009-0472-0182.
25 Society of Automotive Engineers, "IMAC Team 2 - Improved Efficiency, Final
Report," April, 2006. This document is available in Docket EPA-HQ-OAR-2009-0472-
0156.
26 Memo to docket, "Meeting with Delphi and Presentation to EPA," March, 2009. This
document is available in Docket EPA-HQ-OAR-2009-0472-0196.
27 Mathur, G., "Experimental Investigation with Cross Fluted Double-Pipe Suction Line
Heat Exchanger to Enhance AC System Performance," SAE 2009-01-0970, 2009. This
document is available in Docket EPA-HQ-OAR-2009-0472.
28 Society of Automotive Engineers Surface Vehicle Standard J2765, issued October,
2008, http://www.sae.org. This document is available in Docket EPA-HQ-OAR-2009-
0472-0171.
29 Barbat, T. et. AL, "CFD Study of Phase Separators in A/C Automotive Systems," SAE
2003-01-0736, 2003. This document is available in Docket EPA-HQ-OAR-2009-0472-
0187.
30 EPA, "AC Idle Test Results_Summary for docket," August, 2009. This document is
available in Docket EPA-HQ-OAR-2009-0472-0183.
31 "2009 Ward's Automotive Yearbook," Ward's Automotive Group, 2009. ISBN
Number 978-0-9105-89-26-0. This document is available in Docket EPA-HQ-OAR-
2009-0472-0248.
32 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. This document is available in Docket EPA-HQ-OAR-2009-
0472.
2-48
-------
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 rale, in developing the 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 CO2 performance
standards for model year 2016 which would result in the new vehicle fleet in the State of
California having CO2 performance equal to the performance from the California Pavley 1
standards. This technical analysis is documented in this sub-chapter of the RIA.
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
(COi 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.
3.1.2 Overview of Equivalency Calculation.
The calculation of the fleet-wide national MY 2015 and MY 2016 COi 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 COi emissions achieved in CA from
the 2015 and 2016 model year fleets.
2. The estimate of fleetwide average COi emissions was disaggregated into achieved car
and track COi emission levels at the national level using the new car and truck
definitions for this rule.
3-1
-------
Regulatory Impact Analysis
3. Based on the anticipated national fleet mix, the achieved car and track 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 COi 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
COi equivalent emissions due to methane (CfLO and nitrous oxide (N2O), are shown in Table
3-2. CH4 and N2O were excluded because the EPA program separately addresses these
emissions (Preamble section III).
Table 3-2 California Regulatory Standards excluding Cftt 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 tracks 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
3-2
-------
Technical Basis of the Standards
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.
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 COi emissions for the new California fleet. As an example in
2016:
Achieved Fleetwide COi 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 CO2 equivalent tailpipe emission level expected in California in
2016 is 248 grams/mile.
3-3
-------
Regulatory Impact Analysis
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 COi 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).
Table 3-4 Summary of Fleet Description Methods
REGULATOR
National Highway Transit
Safety Association (CAFE
Through MY 2010)
California ARE
EPA
CAR DEFINITION
Car - Passenger Car
Car-PC + LDTl
Passenger Automobile - PC
+ 2 wheel drive SUVs below
6,000 GVW
TRUCK DEFINITION
Truck - LDT1-4 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. 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 CC>2 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
Final Rule
To determine the MY 2015 and MY 2016 fleetwide targets under the EPA final rule,
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
3-4
-------
Technical Basis of the Standards
is expected to be comprised of approximately 66.4% PA and 33.6% NPA. 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 CO2 / mile, while the 2016
target is 250 grams CO2/mile.
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 for the 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 3.1.2.1, the AEO estimates were used to project PC/LDT1
fractions in CA, and PA and NPA sales fractions nationally (Table 3-6).
Table 3-6 National PA and NPA Sales Fractions estimated in March 2009 AEO Projections
Regulatory Class
AEO Car fraction
AEO Truck fraction
PC/LDT1 in CA
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%
One commenter, Yuli Chew, stated that he thought that the 6% per year alternative
was more representative of Pavley I levels. As this analysis shows, the national GHG
standard would provide 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 In their comments on the proposal, the California Air
Resources Board agreed that the standards presented in this rulemaking align with
California's Pavley greenhouse gas emissions standards, and ultimately arrive at the same
stringency as California's standards in MY 2016.
3-5
-------
Regulatory Impact Analysis
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
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 CO2 emission
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
3-6
-------
Technical Basis of the Standards
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 finalizing 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 CO2 standards for
both cars and light trucks: footprint, curb weight, engine displacement, interior 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 CO2 emissions. Since
emissions of CO2 are controlled, there must be a reasonable degree of correlation from a
technical perspective between an attribute and vehicle CO2 emissions performance. 2) The
relationship between the attribute and potential CO2 reducing 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 CO2 control
strategies. 3) How much the attribute would encourage compliance strategies that tend to
circumvent the goal of CO2 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 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 basing the individual manufacturers fleetwide CO2 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
3-7
-------
Regulatory Impact Analysis
2011CA1
curves we
40-
38 -
36-
34-
32
S 30
28
26
24-
22
FE program contains more detailed information how the MY 201 1 car and truck
ire generated.
NHTSA Final MY 2011 Standards for Cars and Trucks
^\
\
\
\
\
x_
50 55 60
Footprint (ft2)
Figure 3-1 NHTSA Reformed CAFE Curve for MY 2011 Cars
3-8
-------
Technical Basis of the Standards
NHTSA Final MY 2011 Standards for Cars and Trucks
40
38
36
34
32
a 30
28
26
24
22
20
35
40
45
50 55 60
Footprint (ft2)
65
70
75
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 CO2 standards based on the
four criteria outlined above.
3.2.1.1 Correlation to Tailpipe CO2 Emissions
Figure 3-3 and Figure 3-4 describe the relationship of tailpipe CO2 emissions and
vehicle footprint. These figures were generated using the manufacturers' 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 do
not appear to have changed.
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
3-9
-------
Regulatory Impact Analysis
should be noted that NHTSA used non-sales-weighted minimum absolute difference (MAD)
regressions to develop the slopes of the fuel economy and CO2 emission standards. The
preamble of this final rule discusses the reasons for use of non-sales-weighted MAD
regressions for this purpose.
500
450 --
400 --
350 --
E
3 300
O
o
250 --
200 --
150 --
y=8.46x-89.866
R2 = .283
100
36
38
40
42
46 48
Footprint (ft2)
50
52
54
56
Figure 3-3 Model Year 2007 Cars; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
Footprint
3-10
-------
Technical Basis of the Standards
750
650 --
550 --
E
3 450
O
o
350 --
250 --
150
y = 4.172x + 170.17
R2=.331
35 40 45 50 55 60 65
Footprint (ft2)
70
75
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 R2 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 CO2 emissions. As vehicle
size increases, its CO2 emissions tend to increase.
3.2.1.2 Relationship with CO2-Reducing Strategies
The footprint attribute would encourage all CO2 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 CO2 emissions. However, smaller vehicles
would have smaller footprints and would be subject to lower, more stringent, CO2 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.
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 CO2 target.
Therefore, in choosing the footprint attribute, the use of lightweight material would remain a
3-11
-------
Regulatory Impact Analysis
viable compliance option, an important factor as lightweight materials can simultaneously
reduce mobile CO2 emissions and improve vehicle safety. NHTSA came to the same
conclusion in its Reformed CAFE rulemaking.7 EPA is assuming that manufacturers can and
will lightweight their vehicles at a given footprint level as a potential compliance strategy.
EPA discusses the relationship of vehicle weight and safety in section 7.6 of this RIA .
3.2.1.3 Sensitivity of CO2 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 CO2 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. 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 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 finalizing this metric in
today's final rule, 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. See Section 3.1. 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 section III.C. 1 of the preamble and Chapter 2 of the RIA, which offer an additional
layer of consistency.
3.2.2 Alternative Attributes
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 CO2 emissions versus curb weight for 2007 car and truck models
respectively, where circle size indicates the sales volume of each model.
3-12
-------
Technical Basis of the Standards
500
450
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
3-13
-------
Regulatory Impact Analysis
750
650 --
550 --
450 --
O
o
350 ---
250 --
150
y=0.0593x + 126.85
R2= .521
3000 3500
4000
4500
5000 5500
Curb Weight (Ibs)
6000
6500
7000
7500
Figure 3-6 Model Year 2007 Trucks; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
Curb Weight
For both cars and tracks, 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 tracks, indicating a substantial relationship of the current fleet's
curb weight and CO2 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.8 In fact, these studies have
shown that a vehicle's size is a more important factor than weight in its effect on occupant
safety. Section 7.6 of this RIA discusses in greater detail EPA's perspective on vehicle weight
and safety. In a weight-based attribute system, a lower weight would correspond to a more
stringent CO2 standard. While this would discourage downsizing as a compliance strategy,
it's important to recognize that weight as an attribute for determining tailpipe CO2 standards
would discourage the use of lightweight materials, even though advanced lightweight
materials could simultaneously reduce CO2 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
3-14
-------
Technical Basis of the Standards
relatively high rate of growth of vehicle weight; it has grown 1.0 - 1.5% per year since the
late 1980s.9 This development would have negative environmental consequences by
increasing overall CO2 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 CO2 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 CO2 emissions and engine
displacement for 2007 cars and trucks, respectively.
500
450 --
400 --
350 --
E
5 300
O
o
250 --
200 --
150 --
100
45.333x + 171.53
R2 = .667
3 4
Engine Displacement (L)
Figure 3-7 Model Year 2007 Cars; Sales-weighted Linear Regression of CO2 Tailpipe Emissions and
Engine Displacement.
3-15
-------
Regulatory Impact Analysis
750
650 --
550 --
450 --
O
o
350 --
250 --
150
y = 45.094x + 217.45
R2=.619
4 5
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 CO2-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-cylinder 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 CO2
emissions (See Chapter 1 of the RIA) 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 CO2 emissions.
3-16
-------
Technical Basis of the Standards
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 CO2 for model year 2007 cars.
MY 2007 Cars: Tailpipe C02 Emissions by Interior Volume
R =0.0036
700
600
500
IN
300
200
i5g
*r«*!5»*: i!? *
*"Kfx}?*
100
50
70
90
110 130
Interior Volume (ft3)
150
170
190
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 CO2 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 CO2 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 finalizing 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 CO2-reducing technologies, there is significant potential for CO2 improvement. Figure
3-10 and Figure 3-11 below illustrate the breakdown of the model year 2007 fleet in terms of
occupant capacity.
3-17
-------
Regulatory Impact Analysis
Model Year 2007 Cars: Percentage of Sales by Occupant Capacity
Figure 3-10 Model Year 2007 Cars; Percentage of Sales by Occupant Capacity
3-18
-------
Technical Basis of the Standards
MY 2007 Cars: Range of Tailpipe CO2 Emissions by Occupant Capacity
ro 400
O
o
I
B- 300
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 CO2 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 finalizing them for that reason. The three remaining attribute options—footprint, curb
weight, and engine displacement—are all reasonable choices in terms of correlation with CO2
emissions levels, with weight having the best correlation to CO2 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 economy10'11 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.
3-19
-------
Regulatory Impact Analysis
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 CO2
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 CO2 standards then the form of the standards would be
compatible with NHTSA's use of footprint in their Reformed CAFE program.
For these reasons, EPA therefore believes that the footprint attribute is the best choice
of the attributes discussed, from both an engineering and public policy standpoint and is using
footprint in the CO2 standard-setting process for this rule.
EPA is implementing the footprint attribute in this CO2 control program via a
piecewise linear function. As mentioned above, this is the equivalent to the shape finalized by
NHTSA for its CAFE standards for model years 2012-2016. The shape of this function with
respect to CO2 is reflected in Figures I.E.3-3 and I.E.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.
3-20
-------
Technical Basis of the Standards
400.0
350.0
1 300.0 -
250.0
200.0
150.0
400.0
350.0
'£ 300.0
o 25°'°
1.50.0
35
2012
2016
35 40 45 50 55 60 65
Footprint (sq. feet)
Figure 3-12 CO2 (g/mi) Car standard curves
40
45
50 55
Footprint (sq. feet)
65
70
Figure 3-13 CO2 (g/mi) Truck standard curves
3-21
-------
Regulatory Impact Analysis
Implementing the CO2 emission standards in this manner provides consistency with
NHTSA's CAFE standards. Section II of the preamble and Chapter 2 of the joint TSD
contain more information on how EPA and NHTSA defined the piecewise linear CO2 target
function.
References
References can be found in EPA docket EPA-HQ-OAR-2009-0472.
1 Energy Information Administration. Annual Energy Outlook 2009.
http://www.eia.doe.gov/oiaf/aeo/index.html
2 NHTSA Model Year 2011 Rule. RIN 2127-AK29. Average Fuel Economy Standards.
Passenger Cars and Light Trucks. Model Year 2011. [Docket No. NHTSA-2009-0062]
3 NHTSA Model Year 2011 Rule. RIN 2127-AK29. Average Fuel Economy Standards.
Passenger Cars and Light Trucks. Model Year 2011. [Docket No. NHTSA-2009-0062]
4 NHTSA Model Year 2011 Rule. RIN 2127-AK29. Average Fuel Economy Standards.
Passenger Cars and Light Trucks. Model Year 2011. [Docket No. NHTSA-2009-0062]
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. ISBN 0-
309-07601-3. 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 (ISBN 0-309-07601-3).
8 71 FR at 17596; 2002 NAS Report at 24 (ISBN 0-309-07601-3).
9 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
10 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
11 Guidelines for Analytically Derived Fuel Economy. March 11, 2004
http://www.epa.gov/otaq/cert/dearmfr/ccd0406.pdf
3-22
-------
Results of Final and Alternative Standards
CHAPTER 4: Results of Final and Alternative Standards
4.1 Introduction
There are many ways for a manufacturer to reduce CC>2 emissions from any given
vehicle. 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 (within the
constraints of sufficient lead time). Thus, for a variety of levels of COi emissions
control, there are an almost infinite number of technology combinations which produce
the desired CO2 reduction. As part of the process of developing the proposed rule, EPA
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. EPA created OMEGA in 2008 and has continued to update its
algorithms through the present. OMEGA underwent a formal peer review process in the
Spring of 2009, and version 1.0 became publicly available in the NPRM docket and on
EPA's web site shortly after publication of the NPRM. The model and a summary of the
peer review process can be found on EPA's web site at:
http://www.epa.gov/otaq/climate/models.htm. EPA continues to use the OMEGA model
here to estimate the technology and cost associated with the final CO2 emission
standards.
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 Joint 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 Regulatory Impact
Analysis (RIA) 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 (see chapter 1 of this RIA) 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
4-1
-------
Regulatory Impact Analysis
Chapter 1 of this RIA and Chapter 3 of the 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 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 Joint TSD.
The fourth type of data describes the CCh emission standards being modeled.
These include the COi emission equivalents of the 2011 MY CAFE standards and the
final CO2 standards for 2016. As described in more detail in Chapter 2 of this RIA 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 COi 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 the 2008 model year vehicles which
comprise our baseline fleet may already 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. While we received no adverse comment on how this process was conducted for the
NPRM, we have improved this process and hopefully made it easier for interested parties
to perform their own analyses in the future.
The first step in the updated process is to breakdown the available GHG control
technologies into five groups: 1) engine-related, 2) transmission-related, 3) hybridization,
4) weight reduction and 5) other. Within each group we gave each individual technology
a ranking which generally followed the degree of complexity, cost and effectiveness of
the technologies within each group. More specifically, the ranking is based on the
premise that a technology on a 2008 baseline vehicle with a lower ranking would be
replaced by one with a higher ranking which was contained in one of the technology
packages which we included in our OMEGA modeling. The corollary of this premise is
that a technology on a 2008 baseline vehicle with a higher ranking would be not be
replaced by one with an equal or lower ranking which was contained in one of the
technology packages which we chose to include in our OMEGA modeling. Table 4-1
presents the technologies and the rankings which we developed for the analyses
4-2
-------
Results of Final and Alternative Standards
supporting the final rule. We do not show any rankings in Table 4-1 for the "other"
technologies, such as improved rolling resistance, electric power steering, etc. These
technologies are assumed to be added to baseline vehicles which do not already have
these technologies.
Table 4-1 Rankings of Individual Technologies
Ranking
1
2
3
4
5
6
7
8
Engine
Intake Cam Phasing
Cylinder Deactivation -
OHV
Dual or Coupled Cam
Phasing
Cylinder Deactivation -
OHC
Variable valve lift
Diesel
Power- Split
2-Mode
Plug-In Electric
Battery Electric
Trans-
mission
CVT
6 Speed
Auto
6 Speed
Manual
Dual
Clutch
Power- Split
2-Mode
Plug-In
Electric
Battery
Electric
Hybrid
42 Volt
Start-Stop
Integrated Motor
Assist
Power- Split
2-Mode
Plug-In Electric
Battery Electric
Weight
3%
5%
10%
Each baseline vehicle was assigned a ranking in each of four categories based on
the maximum ranking of any of its applicable technologies in each technology category.
For example, a vehicle with an OHC engine with both coupled cam phasing and cylinder
deactivation was assigned an engine technology ranking of "3", the ranking applicable to
cylinder deactivation, since its ranking is higher than that for coupled cam phasing. The
same was done for the technology packages. The engine technology for this example
baseline vehicle was left alone whenever a technology package had an engine ranking of
3 or less.
It should be noted that the strong hybrid packages were assigned engine and
transmission rankings, as well as hybrid rankings. The application of strong hybrid
technology affects the type of engine used in the vehicles. For example, it is not
reasonable to add cylinder deactivation or variable valve lift to vehicles which already
have power-split or 2-model hybrid systems.
Two engine-related technologies are not shown in Table 4-1: gasoline direct
injection and turbocharging. Whenever a technology package included gasoline direct
injection, the baseline engine was converted to gasoline direct injection. If the baseline
engine was already of gasoline direct injection design, this aspect of the engine was left
unchanged.
4-3
-------
Regulatory Impact Analysis
The possibility that a baseline engine was already turbocharged was handled
slightly differently in order to maintain the manufacturer's established tendency to
turbocharge its engines. If the engine of a baseline vehicle was turbocharged, this
turbocharging was assumed to continue with the addition of any technology package
which did not include strong hybridization (i.e., power-split, 2-mode, plug-in or battery
electric). In addition, if a package included either cylinder deactivation or variable valve
lift, neither of these technologies was added with the addition of that package. The
turbocharger was assumed to supplant this technology.
In the second step of the process, we used these rankings to estimate the complete
list of technologies which would be present on each baseline vehicle after the application
of each technology package. We then used the EPA lumped parameter model to estimate
the total percentage CO2 emission reduction associated with the technology present on
the baseline vehicle (termed package 0), as well as the total percentage reduction after
application of each package. This process was repeated to determine the total cost of all
of the technology present on the baseline vehicle and after the application of each
applicable technology package.
The third step in this process is to determine the degree of each technology
package's incremental effectiveness and incremental cost is affected by the technology
already present on the baseline vehicle. The degree to which a technology package's
incremental effectiveness is reduced by technology already present on the baseline
vehicle is termed the technology effectiveness basis, or TEB, in the OMEGA model The
value of each vehicle's TEB for each applicable technology package is determined as
follows:
, ( TotalEjfect^ } ( 1-TotalEjfect pl
(l-TotalEjfectvl) (1-TotalEjfect^
TEE, =
1-
1-TotalEjfect pl
1-TotalEjfect^
Where
TotalEffectyj = Total effectiveness of all of the technologies present on the
baseline vehicle after application of technology package i
TotalEffectv,i-i = Total effectiveness of all of the technologies present on the
baseline vehicle after application of technology package i-1
TotalEffectP4 = Total effectiveness of all of the technologies included in
technology package i
TotalEffectp;i-i = Total effectiveness of all of the technologies included in
technology package i-1
4-4
-------
Results of Final and Alternative Standards
The degree to which a technology package's incremental cost is reduced by
technology already present on the baseline vehicle is termed the cost effectiveness basis,
or CEB, in the OMEGA model. The value of each vehicle's CEB for each applicable
technology package is determined as follows:
CEBi = 1 - (TotalCostVji - TotalCostv,i-i) / (TotalCostp4 - TotalCostpj.i)
Where
TotalCosty = total cost of all of the technology present on the vehicle after
addition of package i or i-1 to baseline vehicle v
TotalCostp = total cost of all of the technology included in package i or i-1
i = the technology package being evaluated
i-1 = the previous technology package
The values of CEB and TEB are capped at 1.0 or less, since a vehicle cannot have
more than the entire package already present on it. In other words, the addition of a
technology package cannot increase emissions nor reduce costs. (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 value of a specific TEB or CEB can be negative,
however. This implies that the incremental effectiveness or the incremental cost of
adding a package can be greater than that when adding the packages in sequence to a
vehicle with no baseline technology.
An example of this is a baseline vehicle with a 6 speed manual transmission. All
of our technology package effectiveness and cost estimates are estimated for specified
baseline vehicles, all of which have 4 speed automatic transmissions. Our technology
packages improve this transmission, sometimes to a 6 speed automatic transmission and
then a dual clutch transmission and sometimes directly to a dual clutch transmission.
Subsequent packages may then strongly hybridize the vehicle. If a baseline vehicle has a
6 speed manual transmission, this transmission is unaffected by the technology packages
which include either a 6 speed automatic transmission or a dual clutch transmission, since
the manual transmission is both cheaper and/or more efficient than these other
transmissions. However, when the vehicle is hybridized, this manual transmission is
replaced. The incremental cost of changing this vehicle to a power-split hybrid design,
for example, is greater than that for a vehicle with a dual clutch transmission, since the
credit for removing the manual transmission is less than that for the dual clutch
transmission. The negative CEB causes the OMEGA model to apply a cost for this
power-split package which is slightly higher than that for the typical baseline vehicle.
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
4-5
-------
Regulatory Impact Analysis
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
on vehicles, and thus ensures that the model does not double count technology
effectiveness and cost associated with complying with the 2011 MY CAFE standards and
the final CO2 standards.
Table 4-2 and Table 4-3 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 4-2 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
1 2.7%
21 .2%
1 8.5%
1 7.6%
21 .3%
1 8.2%
1 4.2%
0.2%
1.0%
4.1%
5.3%
1 1 .2%
34.0%
8.5%
0%
1 5.0%
19.1%
21 .7%
26.2%
2
1 6.9%
24.5%
1 9.6%
33.6%
33.5%
41 .4%
1 5.6%
0.2%
0.1%
5.2%
0.8%
1 3.4%
32.0%
32.1%
0%
27.4%
40.8%
1 3.0%
45.0%
3
1.2%
1 2.5%
2.9%
0.0%
5.9%
6.6%
0.2%
-0.9%
-0.5%
0.6%
-4.1%
0.1%
6.5%
0.0%
0%
2.8%
0.3%
1.0%
0.0%
4
-2.3%
-0.7%
0.0%
-3.6%
-0.1%
-0.7%
2.5%
0.2%
0.1%
0.0%
0.9%
0.2%
-0.1%
0.5%
0%
2.0%
3.5%
0.0%
0.0%
5
0.0%
0.0%
0.0%
-0.5%
-0.6%
0.0%
-4.5%
-0.1%
-0.1%
0.0%
0.0%
0.0%
0.3%
0.0%
0%
0.0%
0.0%
0.0%
0.0%
N/A: No such package for that vehicle type
4-6
-------
Results of Final and Alternative Standards
Table 4-3 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
1
1.8%
3.7%
7.9%
4.3%
1 0.2%
3.1%
6.4%
0.2%
0.3%
0.6%
1.4%
1.5%
3.9%
0.0%
0.0%
2.6%
2.3%
1 1 .9%
2.1%
2
30.4%
33.7%
24.5%
36.4%
25.5%
32.4%
27.9%
0.1%
0.1%
4.2%
1.0%
4.6%
1 4.2%
14.1%
0.0%
52.5%
48.3%
27.8%
48.3%
3
1.5%
19.1%
4.0%
0.0%
8.4%
6.4%
0.2%
0.0%
0.0%
0.4%
0.0%
0.0%
7.7%
0.0%
0.0%
3.7%
1.5%
0.0%
0.0%
4
-0.6%
0.0%
0.0%
4.6%
-0.2%
3.7%
1.6%
0.0%
0.0%
0.5%
0.0%
0.0%
-0.5%
0.4%
0.0%
2.6%
4.3%
0.0%
0.0%
5
0.0%
0.0%
0.0%
-3.3%
0.1%
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%
As mentioned above, for the market data input file utilized by OMEGA
characterizing 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 TSD Chapter 3. Because EPA chose to apply technologies in packages,
(a methodology endorsed by many commenters and not challenged by any) and 2008
vehicles are equipped with individual technologies in a wide variety of combinations,
accounting for the presence of specific technologies in terms of their proportion of
package cost and CO2 effectiveness requires careful, detailed analysis. The first step in
this analysis is to develop a list of individual technologies which are either contained in
each technology package, or would supplant the addition of the relevant portion of each
technology package. An example would be a 2008 MY vehicle equipped with variable
valve timing and a 6-speed automatic transmission. The cost and effectiveness of
variable valve timing would be considered to be already present for any technology
packages which included the addition of variable valve timing or technologies which
went beyond this technology in terms of engine related CO2 control efficiency. An
example of a technology which supplants several technologies would be a 2008 MY
vehicle which was equipped with a diesel engine. The effectiveness of this technology
4-7
-------
Regulatory Impact Analysis
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%. 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.
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
4-8
-------
Results of Final and Alternative Standards
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. The development of
these factors is described in detail in the previous sub-section.
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.
C02t_1x(l-CAPxAIE]
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
AvgVehicleCostMFK =
o MrK rj-, t IT^I >o 7
TotalFleetSales MFR
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
the core model calculations, the user must enter into the technology input file the
4-9
-------
Regulatory Impact Analysis
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.
EPA received no adverse comment on this approach and no changes in this
methodology have been made since the NPRM.
4.3 Modeling Process
In order to determine the technology costs associated with this final 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, pursuant to 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 rule. 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 final rule alone, EPA seeks to subtract out any costs associated with
meeting any existing standards related to GHG emissions. EPA consequently ran
OMEGA a second time to calculate the cost of meeting the EPA's final standards in
2016, and then subtracted the results of the reference case model run to determine the
costs of this final 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 final rule allows for averaging across a manufacturer's cars and trucks, the model
determines the CO2 emission standard applicable to each manufacturer's car and truck
sales from the two sets of coefficients describing the piecewise linear standard functions
for cars and trucks in the inputs, and creates a combined car-truck standard. This
combined standard considers the difference in lifetime VMT of cars and trucks, as
indicated (for example) in the final regulations which governs 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 these vehicles'
footprint values. When evaluating the 2011 MY CAFE standards, the car-truck trading
was limited to 1.2 mpg. When evaluating the final CO2 standards, the OMEGA model
was run only for MY 2016. OMEGA is designed to evaluate technology addition over a
complete redesign cycle and 2016 represents the final year of a redesign cycle starting
with the first year of the final CO2 standards, 2012. Estimates of the technology and cost
for the interim model years are developed from the model projections made for 2016.
This process is discussed in Chapter 6 of EPA's RIA to this final rule. When evaluating
the 2016 standards using OMEGA, the final 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 10.2 g/mi of A/C credit in 2016 for each car sold and 11.5 g/mi of A/C credit for
each truck sold. Thus, the sales-weighted 2016 CO2 target for the fleet evaluated using
OMEGA was 261 g/mi instead of 250 g/mi.
4-10
-------
Results of Final and Alternative Standards
The cost of the improved A/C systems required to generate this credit was
estimated separately. This is consistent with the final 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 CO2 the technology will reduce.
One component of the cost of adding a technology is its production cost, as discussed
above. However, it is expected that new vehicle purchasers value improved fuel
economy since it reduces the cost of operating the vehicle. Typical vehicle purchasers
are assumed to value the fuel savings accrued over the period of time which they will
own the vehicle, 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:
pp i
TechCost - £ [dFS{ x VMTi ] x -
,, n^, r£C i=\ (I -Gap)
ManufCostEff = - — - - - - - —
Y [[dc<92]x VMT;. ]x7 — 1 — ,
^U (\-Gap)
Where
4-11
-------
Regulatory Impact Analysis
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,
dCC>2 = Difference in COi 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 rule, but a measure of the private cost,
(i.e., a measure of the vehicle purchaser's willingness to pay more for a vehicle with
higher fuel efficiency). Since vehicle operators pay the full price of fuel, including
taxes, they value fuel costs or savings at this level, and the manufacturers will consider
this when choosing among the technology options.
This definition of manufacturer-based net cost-effectiveness ignores any change
in the residual value of the vehicle due to the additional technology when the vehicle is
five years old. 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 rule, this factor was not included in the determination of
manufacturer-based net cost-effectiveness in the analyses performed in support of this
final rule.
The values of manufacturer-based net cost-effectiveness for specific technologies
will vary from vehicle to vehicle, often substantially. This occurs for three reasons.
First, both the cost and fuel-saving component cost, ownership fuel-savings, and lifetime
COi effectiveness of a specific technology all vary by the type of vehicle or engine to
which it is being applied (e.g., small car versus large truck, or 4-cylinder versus 8-
cylinder engine). Second, the effectiveness of a specific technology often depends on the
presence of other technologies already being used on the vehicle (i.e., the dis-synergies.
Third, the absolute fuel savings and CC>2 reduction of a percentage an incremental
reduction in fuel consumption depends on the COi level of the vehicle prior to adding the
technology. EPA believes this manufacturer-based net cost-effectiveness metric is
appropriate for ranking technology in this final program because it considers
4-12
-------
Results of Final and Alternative Standards
effectiveness values that may vary widely among technology packages when determining
the order of technology addition..
4.4 Modeling of CAA Compliance Flexibilities
EPA's final rule incorporates several compliance flexibilities. See generally
section III.C of the preamble to the final rule. 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 final 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 of this RIA, 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.
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
4-13
-------
Regulatory Impact Analysis
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 final A/C credit system. Neither direct nor indirect
A/C emissions are included in the compliance determination towards the final 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 final CO2
emission standards accordingly. Thus, the effective 2016 final 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 final 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, the OMEGA model published with the NPRM
has been upgraded to directly facilitate the trading of car-truck credits on a total lifetime
CO2 emission basis, consistent with the provisions of the proposed and final CO2 rule.
For example, if a manufacturer over-complies with its applicable CO2 standard for cars
by 3 g/mi,sells 1,000,000 cars, and cars have a lifetime VMT of 195,264 miles, it
generates 585,792 metric tons of CO2 credits. If these credits are used to compensate for
under-compliance towards the truck CO2 standard and truck sales are 500,000, with a
lifetime truck VMT of 225,865 miles, the manufacturer's truck CO2 emission level could
be as much as 5.2 g/mi CO2 above the standard.
to'
Under the final rule, 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 the CAFE 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. This credit decreases to 0.8 mpg in 2016. Car-truck
trading is also allowed under the CAFE program, up to 1.0 mpg in 2011. This car-truck
credit trading limitation increases to 1.5 mpg in 2016. Our reference case is a 2016
vehicle fleet complying with the 2011 CAFE standards. Thus, there is some basis for
utilizing the FFV and car-truck credit limits applicable in 2016. However, as the changes
to the FFV and car-truck credit limits over time are part of EISA itself, and the fuel
economy side of these joint NHTSA-EPA rules implements a provision of EISA, these
changes to the FFV credit and car-truck credit trading can be considered to be part of the
fuel economy regulation being promulgated and not part of the baseline or reference case
existing prior to this rule. We believe that this latter classification is the most appropriate
4-14
-------
Results of Final and Alternative Standards
for this rale analysis. Thus, in our reference case, we limit FFV credits to 1.2 mpg and
car-truck trading is limited to 1.0 mpg.
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-track 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 21.5 mpg if the manufacturer generated sufficient credits from
its car fleet. These two fuel economy levels represent CO2 emission levels of 395 and
413 g/mi, respectively, assuming all the vehicles are fueled with gasoline, a difference of
18 g/mi CO2. If the manufacturer's track 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 370 and 386 g/mi, respectively, a
difference of 16 g/mi CO2. In both cases, the difference in terms of mpg is 1.0.
However, the difference in terms of CO2 emissions decreases as the base fuel economy
increases.
The fact that the same limit in terms of fuel economy translates into differing
limits in terms of CO2 emissions complicates the modeling of CO2 emission compliance
using the OMEGA model. The model currently only accepts a single limit on car-track
trading in terms of g/mi CO2 emissions. However, since the limit of 1.0 mpg on car-
truck trading results in a different limit for each manufacturer, this necessitates a separate
model run for each manufacturer when the trading of credits might approach the 1.0 mpg
limit. Also, the OMEGA model is not yet set up to accept FFV credits in terms of mpg.
Thus, the CO2 standards applicable to those manufacturers expected to utilize FFV
credits must be adjusted outside of the model. (Work is underway to facilitate these
credits within the model, but was not completed in time for this final rale analysis.)
Thus, we adjusted the footprint-based standard for each manufacturer expected to
use FFV credits by the level of CO2 emissions equivalent to the maximum 1.2 mpg FFV
credit. 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).
In addition, for manufacturers expected to pay CAFE fines in lieu of compliance,
we substituted the achieved fuel economy levels from NHTSA's Volpe Model
evaluations of the 2011 CAFE standards for these manufacturers' CAFE standards. The
only manufacturer found to prefer paying fines over compliance was Porsche, and then
only for its cars.
We initially ran the OMEGA model with unlimited trading of car-track credits to
determine the degree of trading which was likely to occur. We then determined the car-
truck trading limit in terms of g/mi CO2 for each manufacturer equivalent to 1.0 mpg and
determined if this limit had been exceeded. Only three manufacturers were found to
exceed the trading limit in the unlimited trading runs, Mitsubishi, Suzuki and Tata. The
OMEGA input and output files using the latest version of the model can be found under
"EPA OMEGA Model" in the docket to this rule.
4-15
-------
Regulatory Impact Analysis
4.5 Manufacturer-Specific Standards and Achieved CO2 Levels
As described in RIA 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 targets calculated for the
final 2016 model year are presented in Error! Reference source not found.Table 4-4.
Table 4-4 2016 Projected Standards by Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
Car
228.4
232.2
238.3
229.2
230.5
222.1
222.2
224.3
221.2
219.4
225.7
206.1
215.5
207.5
249.9
221.1
218.6
225.1
Truck
282.5
295.0
294.3
304.7
315.7
280.6
278.3
289.3
270.8
269.1
294.4
286.9
267.1
271.9
272.5
294.4
292.7
297.7
Production
Weighted
Average0
243.9
265.8
256.1
257.1
270.5
243.7
230.6
235.5
228.4
239.3
245.4
233.0
234.2
218.0
258.8
245.0
231.6
250.1
VMT Weighted
Average*
245.6
268.1
257.9
259.7
273.6
245.7
231.7
237.0
229.4
241.1
247.5
235.7
235.9
219.3
259.6
247.4
233.2
252.5
Production weighted CO2 levels include reductions from A/C improvements and are weighted by
production only.
*VMT weighted CO2 levels include reductions from A/C improvements and are weighted by both
production and VMT for consistency with CO2 standard levels.
The VMT weighted car and truck standards average out to an overall industry
CO2 stringency of 252.5 g/mi. This number is based on sales and lifetime VMT
weightings of the applicable car and truck standards. The 2016 industry combined CO2
level of 250 g/mi presented by President Obama in his announcement on May 19, 2009
was calculated by weighting car and truck CO2 by sales only and did not consider trading
on a lifetime VMT basis. As shown above, when the combined car and truck standards
above are calculated using a sales weighting alone, the industry combined average results
in250.1g/mi.
The majority of manufacturers representing the vast majority of sales in 2016 are
projected to comply with the final 2016 standards with the addition of technology under
4-16
-------
Results of Final and Alternative Standards
the penetration limits described in Section 4.7 below. However, several smaller volume
manufacturers (at least with respect to U.S. sales) are projected to fall short of
compliance. For a more complete discussion of the feasibility of the standards, please see
Section III.D in the preamble. Table 4-5 below contains the projected achieved levels of
CO2 emissions for each manufacturer from the OMEGA model. Overall, these levels are
very similar to those projected in the NPRM.
Table 4-5 Projected Achieved CO2 Levels in 2016
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Overall
Car
236.3
227.9
233.4
218.2
241.3
207.6
214.5
258.6
213.1
218.2
246.3
223.3
223.2
244.1
197.3
212.8
223.5
223.8
Truck
278.7
298.2
298.3
263.0
305.1
302.0
315.6
323.6
335.2
285.6
297.8
264.0
299.8
332.0
317.7
308.6
326.6
302.5
Production
Weighted
Average0
248.5
265.6
257.3
234.4
271.3
242.5
229.8
284.2
234.3
228.1
262.6
239.6
245.2
273.4
216.8
244.0
241.6
250.8
VMT
Weighted
Average*
249.8
268.1
259.6
235.9
273.6
245.7
231.7
286.5
237.0
229.4
264.3
241.1
247.5
276.3
219.3
247.1
243.9
253.5
Production weighted CO2 levels include reductions from A/C
improvements and are weighted by production only.
*VMT weighted CO2 levels include reductions from A/C
improvements and are weighted by both production and VMT for
consistency with CO2 standard levels.
4.6 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
first used OMEGA to calculate costs reflected in the existing CAFE program, which is
4-17
-------
Regulatory Impact Analysis
the reference case for this analysis. 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 4-6 below.
Table 4-6 Incremental Technology Cost of the Reference Case
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Total
Car
$ 346
$ 33
$ 73
$ 68
$ 31
$
$
$ 611
$
$
$ 468
$ 328
$
$ 473
$ 49
$
$ 228
$ 63
Truck
$ 423
$ 116
$ 161
$ 62
$ 181
$
$ 69
$ 1,205
$ 42
$
$ 683
$ 246
$ 61
$ 706
$ 232
$
$ 482
$ 138
Combined
$ 368
$ 77
$ 106
$ 66
$ 102
$
$ 10
$ 845
$ 7
$
$ 536
$ 295
$ 18
$ 550
$ 79
$
$ 272
$ 89
EPA then used OMEGA to calculate the costs of meeting the final 2016
standards, which are displayed in Table 4-7 below, and two alternative scenarios for
sensitivity. In Table 4-7 and Table 4-17, 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.
4-18
-------
Results of Final and Alternative Standards
Table 4-7 Incremental Technology Cost of the Final 2016 CO2 Standards
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Total
Car
$ 1,558
$ 1,129
$ 1,108
$ 962
$ 899
$ 635
$ 802
$ 1,181
$ 667
$ 855
$ 1,536
$ 817
$ 686
$ 1,506
$ 1,015
$ 381
$ 1,848
$ 870
Truck
$ 1,195
$ 1,501
$ 1,442
$ 790
$ 1,581
$ 473
$ 425
$ 680
$ 247
$ 537
$ 931
$ 1,218
$ 1,119
$ 759
$ 537
$ 609
$ 972
$ 1,099
Combined
$ 1,453
$ 1,329
$ 1,231
$ 899
$ 1,219
$ 575
$ 745
$ 984
$ 594
$ 808
$ 1,343
$ 978
$ 810
$ 1,257
$ 937
$ 455
$ 1,694
$ 948
EPA estimates that the additional technology required for manufacturers to meet
the GHG standards for this final rule will cost on average $948/vehicle. This cost is
roughly $100 lower than that projected in the NPRM. This difference is due primarily to
a reduction in the estimated cost for the various technologies being added to the vehicles.
4.7 Technology Penetration
The major technologies chosen by OMEGA are described in the Table 4-8
through Table 4-12 for the reference case and in Tables 4-11 through 4-13 for the control
case for cars, trucks, and combined fleets. 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
Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
53%
0%
23%
0%
6%
0%
0%
0%
11%
45%
0%
88%
0%
80%
85%
7%
87%
12%
DEAC-
OHC
10%
0%
20%
0%
0%
9%
0%
0%
0%
0%
0%
0%
0%
0%
64%
0%
3%
2%
Turbo
43%
1%
5%
4%
3%
0%
0%
0%
11%
3%
0%
88%
12%
0%
0%
0%
84%
8%
Diesel
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
6SPD
Auto
40%
5%
47%
32%
14%
0%
0%
0%
15%
25%
0%
0%
0%
0%
34%
21%
13%
15%
DCT
45%
0%
32%
0%
0%
0%
0%
0%
0%
50%
0%
41%
0%
80%
64%
0%
77%
8%
42 V S-S
12%
0%
21%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
80%
64%
0%
4%
2%
IMA
0%
0%
0%
0%
0%
3%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Power
Split
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
1%
0%
0%
0%
0%
15%
0%
3%
2-Mode
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
% Weight
Reduction
1 .8%
0.0%
1.1%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.6%
0.0%
2.8%
0.0%
4.0%
3.8%
0.0%
2.4%
0.3%
4-20
-------
Results of Final and Alternative Standards
Table 4-9 2016 Technology Penetration in the Reference Case-Trucks
Manufacturer
BMW
Chrysler
Daimler
Ford
General
Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
20%
0%
24%
1%
0%
4%
0%
0%
26%
13%
0%
100%
0%
18%
85%
7%
99%
7%
DEAC-
OHC
16%
0%
24%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
37%
0%
17%
2%
Turbo
0%
0%
16%
0%
0%
4%
0%
0%
26%
0%
0%
50%
3%
0%
51%
0%
69%
3%
Diesel
0%
0%
16%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
1%
0%
6SPD
Auto
84%
28%
62%
19%
17%
0%
23%
0%
48%
25%
0%
15%
0%
18%
15%
16%
15%
18%
DCT
16%
0%
38%
0%
0%
0%
0%
0%
0%
13%
0%
84%
0%
0%
85%
0%
85%
4%
42 V S-S
16%
0%
38%
0%
0%
0%
0%
0%
0%
0%
0%
85%
0%
0%
85%
0%
85%
4%
IMA
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%
6%
0%
1%
2-Mode
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
% Weight
Reduction
1 .6%
0.0%
3.8%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.4%
0.0%
6.6%
0.0%
0.5%
8.5%
0.0%
5.1%
0.3%
4-21
-------
Regulatory Impact Analysis
Table 4-10 2016 Technology Penetration in the Reference Case - Combined Cars and Trucks
Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
44%
0%
23%
0%
3%
2%
0%
0%
13%
32%
0%
92%
0%
70%
85%
7%
89%
10%
DEAC-
OHC
12%
0%
22%
0%
0%
6%
0%
0%
0%
0%
0%
0%
0%
0%
54%
0%
5%
2%
Turbo
30%
0%
8%
3%
1%
2%
0%
0%
13%
2%
0%
75%
9%
0%
20%
0%
81%
7%
Diesel
0%
0%
6%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0.2%
6SPD
Auto
53%
18%
52%
27%
15%
0%
3%
0%
20%
25%
0%
5%
0%
3%
27%
19%
14%
16%
DCT
37%
0%
34%
0%
0%
0%
0%
0%
0%
35%
0%
55%
0%
67%
73%
0%
78%
7%
42 V S-S
13%
0%
26%
0%
0%
0%
0%
0%
0%
0%
0%
38%
0%
67%
73%
0%
18%
3%
IMA
0%
0%
0%
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0.2%
Power
Split
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
1%
0%
0%
0%
0%
12%
0%
2.5%
2-Mode
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0.0%
% Weight
Reduction
1.7%
0.0%
2.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.1%
0.0%
4.1%
0.0%
3.4%
5.7%
0.0%
2.8%
0.3%
4-22
-------
Results of Final and Alternative Standards
Table 4-11 2016 Technology Penetration in the Control Case-Cars
Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
78%
85%
78%
85%
56%
59%
61%
39%
59%
67%
62%
75%
82%
80%
85%
22%
79%
58%
DEAC-
OHC
19%
1%
28%
12%
3%
9%
0%
0%
0%
0%
0%
15%
0%
0%
64%
2%
16%
6%
Turbo
62%
7%
54%
15%
13%
0%
1%
1%
12%
7%
3%
73%
12%
0%
17%
4%
74%
14%
Diesel
8%
0%
6%
0%
0%
0%
0%
0%
1%
1%
0%
12%
0%
0%
0%
0%
13%
1%
6SPD
Auto
13%
30%
10%
32%
4%
0%
0%
0%
11%
21%
1%
0%
0%
0%
14%
15%
10%
11%
DCT
61%
55%
72%
56%
53%
70%
61%
63%
48%
67%
57%
34%
79%
80%
70%
49%
67%
58%
42 V S-S
63%
56%
65%
57%
53%
21%
38%
4%
47%
67%
52%
58%
58%
80%
70%
2%
57%
39%
IMA
0%
0%
0%
0%
0%
3%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Power
Split
4%
0%
3%
0%
0%
0%
0%
0%
0%
0%
1%
0%
0%
0%
4%
15%
2%
4%
2-Mode
10%
0%
10%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
0%
11%
0%
13%
2%
% Weight
Reduction
4.7%
6.2%
4.8%
4.9%
4.7%
3.6%
3.4%
2.2%
3.9%
6.3%
4.8%
3.5%
4.2%
4.0%
4.3%
1.7%
3.9%
3.8%
4-23
-------
Regulatory Impact Analysis
Table 4-12 2016 Technology Penetration in the Control Case-Trucks
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
86%
74%
72%
83%
80%
16%
44%
1%
65%
85%
76%
100%
21%
61%
85%
34%
99%
62%
DEAC-OHC
28%
25%
35%
35%
50%
0%
0%
0%
0%
0%
23%
15%
0%
0%
40%
17%
29%
27%
Turbo
57%
25%
52%
25%
16%
4%
0%
0%
27%
72%
29%
42%
3%
0%
42%
0%
56%
18%
Diesel
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
6SPD
Auto
15%
32%
15%
18%
12%
0%
56%
0%
48%
4%
3%
15%
0%
61%
15%
9%
15%
13%
DCT
69%
50%
70%
68%
70%
12%
0%
1%
39%
85%
72%
69%
21%
0%
70%
20%
70%
49%
42 V S-S
70%
53%
70%
68%
70%
12%
0%
1%
7%
85%
73%
70%
21%
0%
70%
17%
70%
48%
IMA
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%
6%
6%
0%
1%
2-Mode
15%
0%
15%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
0%
9%
0%
15%
1%
% Weight
Reduction
4.1%
5.7%
5.1%
6.6%
6.5%
1 .2%
1 .3%
0.1%
1 .7%
5.9%
6.4%
5.4%
2.1%
1 .8%
7.0%
2.0%
4.2%
4.5%
4-24
-------
Results of Final and Alternative Standards
Table 4-13 2016 Technology Penetration in the Control Case - Combined Cars and Trucks
Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
80%
79%
76%
84%
67%
43%
59%
33%
60%
74%
66%
83%
60%
77%
85%
26%
82%
60%
DEAC-
OHC
21%
13%
30%
21%
25%
6%
0%
0%
0%
0%
7%
15%
0%
0%
55%
7%
18%
13%
Turbo
61%
17%
53%
19%
14%
2%
1%
1%
14%
33%
11%
62%
9%
0%
27%
3%
71%
15%
Diesel
6%
0%
5%
0%
0%
0%
0%
0%
1%
0%
0%
8%
0%
0%
0%
0%
11%
0.9%
6SPD
Auto
13%
31%
12%
27%
8%
0%
8%
0%
17%
14%
2%
5%
0%
10%
14%
13%
10%
12%
DCT
63%
52%
72%
60%
61%
49%
52%
52%
47%
74%
62%
45%
58%
67%
70%
40%
68%
55%
42V
S-S
65%
54%
67%
61%
61%
18%
32%
4%
41%
74%
58%
62%
44%
67%
70%
7%
60%
42%
IMA
0%
0%
0%
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Power
Split
3%
0%
2%
0%
0%
0%
0%
0%
0%
0%
1%
0%
0%
0%
4%
12%
1%
3%
2-Mode
12%
0%
12%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
0%
11%
0%
14%
1%
% Weight
Reduction
4.5%
5.9%
4.9%
5.5%
5.5%
2.7%
3.1%
1 .8%
3.5%
6.2%
5.3%
4.1%
3.4%
3.6%
5.4%
1 .8%
3.9%
4.1%
4-25
-------
Regulatory Impact Analysis
As can be seen, the overall reduction in vehicle weight is projected to be 4.3%. As there
has been a concern in the past that weight reductions are associated with increased safety risk, a
more specific breakdown of the projected weight reduction by vehicle class and weight range is
provided below. For cars below 2950 pounds curb weight, the estimated reduction is 2.3% (62
pounds), while it was estimated to be 4.4% (154 pounds) for cars above 2950 curb weight. For
trucks below 3850 pounds curb weight, the projected 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 estimated 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 4-14.
Table 4-14 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
Average
Weight
Reduction
75 Ibs
153 Ibs
163 Ibs
240 Ibs
186 Ibs
376 Ibs
% Weight
Reduction
2.8%
4.3%
4.7%
5.1%
4.4%
7.0%
4.8 Alternative Program Stringencies
EPA also analyzed the technology cost of two alternative stringency scenarios: 4%/year
and 6%/year. The manufacturers's CO2 targets and achieved levels for standards with these
alternative stringincies are presented in Table 4-15 and Table 4-16 below.
4-26
-------
Results of Final and Alternative Standards
Table 4-15 2016 Standards by Manufacturer in the 4% Sensitivity Case
BMW
Chrysler
Ford
Subaru
General
Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Overall
Achieved CO2 Levels
Car
236.3
228.7
237.3
218.2
245.6
211.1
216.2
258.6
217.7
222.9
246.3
227.5
229.2
244.1
197.3
217.5
223.5
227.3
Truck
278.7
305.2
302.7
271.7
308.1
305.3
329.3
323.6
335.7
285.6
297.8
268.4
299.8
332.0
334.3
311.6
326.6
305.9
Combined
249.8
272.5
263.7
239.4
277.3
249.1
235.5
286.5
240.7
233.3
264.3
245.3
251.6
276.3
222.4
251.2
243.9
256.9
CO2 Standards
Car
232.1
235.9
242.0
233.0
234.2
225.8
225.9
228.0
224.9
223.1
229.4
209.8
219.2
211.3
253.6
224.8
222.3
229.2
Truck
287.4
299.9
299.3
309.6
320.6
285.5
283.3
294.3
275.7
274.0
299.3
291.8
272.1
276.9
277.4
299.3
297.7
299.7
Combined
249.7
272.5
265.1
263.3
278.0
249.9
235.7
256.4
234.8
231.5
253.9
245.6
236.0
235.3
258.0
251.5
237.2
255.8
Table 4-16 2016 Standards by Manufacturer in the 6% Sensitivity Case
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Achieved CO2 Levels
Car
236.3
210.7
214.2
207.6
213.6
194.7
202.9
258.6
189.4
203.9
246.3
212.2
200.2
244.1
186.8
Truck
278.7
273.9
285.0
227.8
290.9
270.9
260.9
323.6
335.2
243.8
297.8
260.6
286.8
332.0
260.6
Combined
249.8
246.9
242.7
215.6
252.7
225.5
212.8
286.5
217.9
210.5
264.3
233.4
227.7
276.3
200.3
CO2 Standards
Car
210.4
214.2
220.3
211.3
212.5
204.1
204.2
206.3
203.2
201.4
207.8
188.1
197.5
189.6
231.9
Truck
258.8
271.3
270.7
281.0
292.1
256.9
254.7
265.7
247.1
245.4
270.7
263.2
243.5
248.3
248.8
Combined
225.8
246.9
240.6
238.9
252.8
225.5
212.8
231.8
211.8
208.7
229.8
220.9
212.1
211.1
235.0
4-27
-------
Regulatory Impact Analysis
Toyota
Volkswagen
Overall
192.9
223.5
206.6
288.3
326.6
284.9
227.1
243.9
236.1
203.1
200.6
207.5
270.7
269.1
271.2
227.4
214.2
231.5
With the reference case the same as that described above in Section 4.1, the costs of the
two alternative control cases are presented in Tables 4-17 and 4-18, respectively, and the
technology penetrations are presented in Table 4-17 through Table 4-24, below.
Table 4-17 2016 Technology Cost in the 4% sensitivity case
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Total
Car
$ 1,558
$ 1,111
$ 1,013
$ 962
$ 834
$ 598
$ 769
$ 1,181
$ 588
$ 766
$ 1,536
$ 733
$ 572
$ 1,506
$ 1,015
$ 323
$ 1,848
$ 811
Truck
$ 1,195
$ 1,236
$ 1,358
$ 616
$ 1,501
$ 411
$ 202
$ 680
$ 238
$ 537
$ 931
$ 1,164
$ 1,119
$ 759
$ 179
$ 560
$ 972
$ 1,020
Combined
$ 1,453
$ 1,178
$ 1,140
$ 836
$ 1,148
$ 529
$ 684
$ 984
$ 527
$ 733
$ 1,343
$ 906
$ 729
$ 1,257
$ 879
$ 400
$ 1,694
$ 883
4-28
-------
Results of Final and Alternative Standards
Table 4-18 2016 Technology Cost in the 6% sensitivity case
BMW
Chrysler
Ford
Subaru
General Motors
Honda
Hyundai
Tata
Kia
Mazda
Daimler
Mitsubishi
Nissan
Porsche
Suzuki
Toyota
Volkswagen
Total
Car
$ 1,558
$ 1,447
$ 1,839
$ 1,173
$ 1,728
$ 894
$ 1,052
$ 1,181
$ 1,132
$ 1,093
$ 1,536
$ 1,224
$ 1,151
$ 1,506
$ 1,426
$ 747
$ 1,848
$ 1,296
Truck
$ 1,195
$ 2,156
$ 2,090
$ 1,316
$ 2,030
$ 891
$ 1,251
$ 680
$ 247
$ 1,083
$ 931
$ 1,840
$ 1,693
$ 759
$ 1,352
$ 906
$ 972
$ 1,538
Combined
$ 1,453
$ 1,827
$ 1,932
$ 1,225
$ 1,870
$ 893
$ 1,082
$ 984
$ 979
$ 1,092
$ 1,343
$ 1,471
$ 1,306
$ 1,257
$ 1,414
$ 799
$ 1,694
$ 1,379
4-29
-------
Regulatory Impact Analysis
Table 4-19 2016 Technology Penetration in the 4% sensitivity case- Cars
Manufacturer
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
78%
85%
78%
73%
57%
48%
61%
45%
82%
85%
65%
75%
82%
80%
85%
9%
79%
55%
DEAC-
OHC
19%
0%
28%
8%
3%
9%
0%
0%
0%
0%
0%
15%
0%
0%
64%
2%
16%
5%
Turbo
62%
7%
54%
15%
6%
0%
1%
1%
12%
3%
3%
73%
12%
0%
17%
0%
74%
12%
Diesel
8%
0%
6%
0%
0%
0%
0%
0%
1%
0%
0%
12%
0%
0%
0%
0%
13%
1%
6SPD
Auto
13%
30%
10%
26%
11%
0%
0%
0%
11%
25%
1%
0%
0%
0%
14%
18%
10%
12%
DCT
61%
55%
72%
56%
46%
59%
61%
68%
71%
63%
60%
34%
79%
80%
70%
35%
67%
54%
42 V S-S
63%
55%
65%
57%
46%
21%
33%
0%
69%
64%
56%
58%
58%
80%
70%
2%
57%
38%
IMA
0%
0%
0%
0%
0%
3%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Power
Split
4%
0%
3%
0%
0%
0%
0%
0%
0%
0%
1%
0%
0%
0%
4%
15%
2%
4%
2-Mode
10%
0%
10%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
0%
11%
0%
13%
2%
% Weight
Reduction
4.7%
6.1%
4.8%
4.6%
4.2%
3.2%
3.2%
2.4%
6.1%
6.6%
5.2%
3.5%
4.2%
4.0%
4.3%
1.1%
3.9%
3.6%
4-30
-------
Results of Final and Alternative Standards
Table 4-20 2016 Technology Penetration in the 4% sensitivity case- Trucks
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
86%
52%
72%
83%
69%
37%
0%
1%
65%
85%
76%
100%
54%
18%
85%
27%
99%
58%
DEAC-
OHC
28%
24%
35%
35%
49%
0%
0%
0%
0%
0%
23%
15%
0%
0%
40%
17%
29%
27%
Turbo
57%
25%
52%
19%
16%
4%
0%
0%
27%
72%
29%
42%
3%
0%
42%
0%
56%
17%
Diesel
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
6SPD
Auto
15%
24%
15%
24%
4%
0%
23%
0%
48%
4%
3%
15%
0%
18%
15%
2%
15%
10%
DCT
69%
50%
70%
62%
69%
32%
0%
1%
39%
85%
72%
69%
54%
0%
70%
20%
70%
50%
42 V S-S
70%
52%
70%
62%
69%
4%
0%
0%
7%
85%
73%
70%
0%
0%
70%
17%
70%
46%
IMA
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%
6%
6%
0%
1%
2-Mode
15%
0%
15%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
0%
9%
0%
15%
1%
% Weight
Reduction
4.1%
5.0%
5.1%
6.2%
6.1%
1 .2%
0.0%
0.0%
1.7%
5.9%
6.4%
5.4%
1.6%
0.5%
7.0%
1 .8%
4.2%
4.2%
4-31
-------
Regulatory Impact Analysis
Table 4-21 2016 Technology Penetration in the 4% sensitivity case - Cars and Trucks Combined
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
80%
67%
76%
77%
62%
44%
52%
37%
79%
85%
69%
83%
72%
70%
85%
15%
82%
56%
DEAC-
OHC
21%
13%
30%
18%
24%
6%
0%
0%
0%
0%
7%
15%
0%
0%
55%
7%
18%
13%
Turbo
61%
17%
53%
16%
11%
2%
1%
1%
14%
31%
11%
62%
9%
0%
27%
0%
71%
14%
Diesel
6%
0%
5%
0%
0%
0%
0%
0%
1%
0%
0%
8%
0%
0%
0%
0%
11%
0.8%
6
SPD
Auto
13%
26%
12%
25%
7%
0%
3%
0%
17%
16%
2%
5%
0%
3%
14%
13%
10%
11%
DCT
63%
52%
72%
58%
57%
49%
52%
57%
66%
72%
64%
45%
70%
67%
70%
30%
68%
53%
42 V S-S
65%
54%
67%
59%
57%
15%
28%
0%
60%
72%
61%
62%
37%
67%
70%
7%
60%
41%
IMA
0%
0%
0%
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Power
Split
3%
0%
2%
0%
0%
0%
0%
0%
0%
0%
1%
0%
0%
0%
4%
12%
1%
3%
2-
Mode
12%
0%
12%
0%
0%
0%
0%
0%
0%
0%
0%
15%
0%
0%
11%
0%
14%
1%
PHEV/EV
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%
0.0%
0.0%
0.0%
0.0%
MS1
0%
14%
0%
18%
5%
34%
24%
43%
9%
13%
7%
0%
35%
3%
0%
23%
0%
16%
MS2
51%
6%
45%
25%
14%
0%
17%
14%
17%
26%
16%
57%
29%
67%
32%
2%
61%
17%
MS3
20%
48%
26%
34%
42%
15%
11%
0%
44%
46%
45%
13%
8%
0%
38%
6%
9%
25%
% Weight
Reduction
4.5%
5.5%
4.9%
5.2%
5.1%
2.5%
2.7%
2.0%
5.5%
6.3%
5.5%
4.1%
3.3%
3.4%
5.4%
1 .3%
3.9%
3.9%
4-32
-------
Results of Final and Alternative Standards
Table 4-22 2016 Technology Penetration in the 6% Sensitivity Case - Cars
BMW
Chrysler
Daimler
Ford
General
Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
78%
85%
78%
85%
85%
72%
70%
75%
85%
84%
84%
75%
83%
85%
85%
71%
79%
79%
DEAC
-OHC
19%
4%
28%
11%
6%
9%
0%
0%
0%
1%
0%
15%
0%
0%
64%
2%
16%
7%
Turb
0
62%
37%
54%
55%
50%
0%
1%
1%
14%
28%
37%
73%
12%
85%
17%
4%
74%
30%
Diese
I
8%
0%
6%
0%
0%
0%
0%
0%
1%
1%
1%
12%
1%
0%
0%
0%
13%
1%
6
SPD
Auto
13%
1%
10%
5%
2%
0%
9%
0%
3%
4%
0%
0%
2%
0%
14%
15%
10%
6%
DCT
61%
84%
72%
74%
83%
70%
61%
75%
80%
78%
80%
34%
79%
85%
70%
56%
67%
71%
42V
S-S
63%
85%
65%
75%
84%
70%
61%
74%
83%
78%
83%
58%
80%
85%
70%
53%
57%
70%
IMA
0%
0%
0%
0%
0%
3%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Powe
r Split
4%
0%
3%
5%
0%
0%
0%
0%
0%
6%
1%
0%
0%
0%
4%
15%
2%
4%
2-
Mode
10%
0%
10%
5%
1%
0%
0%
0%
0%
1%
0%
15%
0%
0%
11%
0%
13%
2%
PHEV/E
V
0.2%
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%
MS1
0%
0%
0%
0%
0%
0%
9%
1%
0%
0%
0%
0%
2%
0%
0%
3%
0%
1%
MS2
48%
2%
47%
12%
5%
13%
21%
24%
29%
3%
4%
70%
46%
0%
53%
39%
63%
23%
MS3
23%
83%
24%
63%
79%
57%
40%
51%
55%
76%
80%
0%
36%
85%
17%
14%
7%
49%
%
Weight
Reductio
n
4.7%
8.4%
4.8%
6.9%
8.2%
6.3%
5.3%
6.3%
6.9%
7.7%
8.2%
3.5%
5.9%
8.5%
4.3%
3.4%
3.9%
6.0%
4-33
-------
Regulatory Impact Analysis
Table 4-23 2016 Technology Penetration in the 6% Sensitivity Case-Trucks
BMW
Chrysler
Daimler
Ford
General
Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswage
n
Fleet
SGDI
86%
85%
72%
85%
85%
61%
85%
1%
89%
85%
85%
100%
85%
85%
85%
70%
99%
78%
DEAC
-OHC
28%
21%
35%
16%
46%
0%
9%
0%
0%
9%
29%
15%
0%
0%
40%
17%
29%
23%
Turbo
57%
61%
52%
60%
34%
28%
76%
0%
45%
62%
39%
42%
28%
85%
42%
7%
56%
37%
Diese
I
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
6
SP
D
Aut
0
15%
4%
15%
3%
2%
4%
12%
0%
13%
4%
0%
15%
6%
0%
15%
32%
15%
10%
DCT
69%
79%
70%
76%
82%
56%
76%
1%
80%
70%
74%
69%
79%
85%
70%
34%
70%
66%
42
VS-
S
70%
82%
70%
76%
82%
56%
76%
1%
80%
70%
74%
70%
79%
85%
70%
34%
70%
66%
IMA
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Powe
r Split
0%
0%
0%
4%
0%
0%
0%
0%
0%
6%
4%
0%
0%
0%
6%
6%
0%
2%
2-
Mode
15%
3%
15%
5%
3%
0%
0%
0%
0%
9%
6%
15%
0%
0%
9%
0%
15%
4%
PHEV/E
V
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%
MS1
0%
0%
0%
0%
0%
4%
0%
0%
5%
0%
0%
0%
6%
0%
0%
29%
0%
6%
MS
2
57%
5%
38%
11%
12%
24%
85%
0%
41%
42%
24%
31%
25%
67%
0%
0%
56%
16%
MS3
13%
77%
32%
65%
70%
32%
0%
1%
39%
28%
50%
39%
54%
18%
70%
34%
14%
50%
% Weight
Reductio
n
4.1%
7.9%
5.1%
7.0%
7.6%
4.5%
4.3%
0.1%
6.1%
4.9%
6.2%
5.4%
6.8%
5.1%
7.0%
4.3%
4.2%
6.0%
4-34
-------
Results of Final and Alternative Standards
Table 4-24 2016 Technology Penetration in the 6% Sensitivity Case - Cars and Trucks combined
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Fleet
SGDI
80%
85%
76%
85%
85%
68%
73%
62%
85%
85%
85%
83%
84%
85%
85%
71%
82%
79%
DEAC-
OHC
21%
13%
30%
13%
25%
6%
1%
0%
0%
4%
8%
15%
0%
0%
55%
7%
18%
12%
Turbo
61%
50%
53%
57%
43%
10%
12%
1%
19%
42%
38%
62%
18%
85%
27%
5%
71%
33%
Diesel
6%
0.05%
5%
0%
0%
0%
0%
0%
1%
0%
0%
8%
1%
0%
0%
0%
11%
0.9%
6
SPD
Auto
13%
3%
12%
4%
2%
1%
9%
0%
4%
4%
0%
5%
3%
0%
14%
20%
10%
7%
DCT
63%
82%
72%
74%
83%
65%
64%
62%
80%
75%
78%
45%
79%
85%
70%
49%
68%
69%
42
VS-
S
65%
83%
67%
75%
83%
65%
64%
61%
82%
75%
81%
62%
80%
85%
70%
47%
60%
69%
IMA
0%
0%
0%
0%
0%
2%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Power
Split
3%
0%
2%
5%
0%
0%
0%
0%
0%
6%
2%
0%
0%
0%
4%
12%
1%
4%
2-
Mode
12%
2%
12%
5%
2%
0%
0%
0%
0%
4%
2%
15%
0%
0%
11%
0%
14%
3%
PHEV/EV
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%
0.0%
0.0%
0.0%
0.0%
MS1
0%
0%
0%
0%
0%
1%
8%
1%
1%
0%
0%
0%
4%
0%
0%
12%
0%
3%
MS2
51%
4%
45%
12%
8%
17%
31%
20%
31%
18%
10%
57%
38%
11%
32%
26%
61%
21%
MS3
20%
80%
26%
64%
75%
48%
34%
42%
52%
57%
71%
13%
42%
74%
38%
21%
9%
49%
% Weight
Reduction
4.5%
8.2%
4.9%
6.9%
7.9%
5.7%
5.2%
5.2%
6.8%
6.6%
7.6%
4.1%
6.3%
8.0%
5.4%
3.7%
3.9%
6.0%
4-35
-------
Regulatory Impact Analysis
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 final 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 CO2 emission levels.
Manufacturers' average COi 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 4-25 below.
Table 4-25 Penetration of Technology in 2008 Vehicles with 2008 Sales: Cars and Trucks
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
SGDI
7.50%
0.00%
0.00%
0.40%
3.10%
1 .40%
0.00%
0.00%
13.60%
0.00%
0.00%
58.60%
0.00%
0.00%
0.00%
6.80%
50.60%
3.80%
DEAC-
OHC
0.00%
0.00%
0.00%
0.00%
0.00%
7.10%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.80%
Turbo
6.10%
0.50%
6.50%
2.20%
1 .40%
1 .40%
0.00%
0.00%
13.60%
0.00%
0.00%
14.90%
9.80%
0.00%
17.30%
0.00%
39.50%
2.60%
Diesel
0.00%
0.10%
5.60%
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%
0.10%
6
Speed
Auto
Trans
86.00%
14.00%
76.00%
29.00%
15.00%
0.00%
3.00%
0.00%
26.00%
10.00%
0.00%
49.00%
0.00%
0.00%
99.00%
21 .00%
69.00%
19.10%
DCT
0.90%
0.00%
7.50%
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%
13.10%
0.50%
42V
S-S
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%
Hybrid
0.00%
0.00%
0.00%
0.00%
0.30%
2.10%
0.00%
0.00%
0.00%
0.00%
0.80%
0.00%
0.00%
0.00%
0.00%
1 1 .60%
0.00%
2.20%
PHEV/EV
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%
0.00%
0.00%
0.00%
0.00%
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 final 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
4-36
-------
Results of Final and Alternative Standards
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
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 CC>2 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" COi levels and their required COi emission level under the final
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.
4-37
-------
Regulatory Impact Analysis
120
100
80
I 20
O
-40
-60
1
0°
Figure 4-1 CO2 Emissions Relative to Fleet Adjusted for Technology and Footprint
The manufacturers projected in Table 4-25 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 g/mi greater COi
emissions than the average 2008 MY truck after accounting for differences in the use of fuel
saving technology and footprint. The lowest adjustment is for Subaru's trucks, which have 50
g/mi CO2 lower emissions than the average truck.
While this comparison confirms the differences in the technology penetrations shown in
Table 4-25, 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 final COi 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 COi emissions over the two-cycle test procedure.
So manufacturers with higher average performance levels will tend to have higher average
emissions for any given footprint. Table 4-26 shows each manufacturer's average ratios of
weight to footprint and horsepower to weight.
4-38
-------
Results of Final and Alternative Standards
Table 4-26 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" CC>2 emissions
was estimated. First, the "no technology" CCh emissions levels were statistically analyzed to
determine the average impact of weight and the ratio of horsepower to weight on CC>2 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 4-27.
4-39
-------
Regulatory Impact Analysis
Table 4-27 Effect of Weight and Performance on "No Technology" Vehicle CO2
Car
Truck
Intercept
(g/mi CO2)
-45.8
-21
Effect of
weight
(g/mi
CO2/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'
emissions for cars and over 70 percent for trucks. These relationships were then used to adjust
each vehicle's "no technology" CC>2 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" CO2 levels and their required CO2 emission level under
the final 2016 standards and subtracted the difference for the industry as a whole. The results are
shown in Figure 4-2.
4-40
-------
Results of Final and Alternative Standards
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' COi 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 CO2 emissions. Tata's emissions are very
close to the industry average. Daimler's vehicles are no more than 10 g/mi above the average for
the industry. This analysis indicates that the primary reasons for the differences in technology
penetrations shown for the various manufacturers in Table 4-27 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 COi
emissions is clear. Producing relatively heavy or high performance vehicles increases CO2
emissions and will require greater levels of technology in order to meet the final CC>2 standards.
4-41
-------
Emissions Impacts
CHAPTER 5: Emissions Impacts
5.1 Overview
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. The primary GHGs of concern are
carbon dioxide (COi), methane, nitrous oxide, hydrofluorocarbons, perfluorocarbons, and
sulfur hexafluoride.1 Mobile sources emitted 31 percent of all U.S. GHG in 2007
(transportation sources, which do not include certain off-highway sources, account for 28
percent) and have been the fastest-growing source of U.S. GHG since 1990.2 Mobile sources
addressed in the recent endangerment finding under CAA section 202(a)—light-duty vehicles,
heavy-duty trucks, buses, and motorcycles-accounted for 23 percent of all U.S. GHG in
2007.3 Light-duty vehicles emit four GHGs-CO2 , methane, nitrous oxide, and
hydrofluorocarbons-and are responsible for nearly 60 percent of all mobile source GHGs and
over 70 percent of Section 202(a) mobile source GHG. For light-duty vehicles in 2007, COi
emissions represent about 94 percent of all greenhouse emissions (including HFCs), and the
CC>2 emissions measured over the EPA tests used for fuel economy compliance represent
about 90 percent of total light-duty vehicle greenhouse gas emissions.4'5
Today's rule quantifies anticipated impacts from the EPA vehicle COi emission
standards. The emissions from the GHGs carbon dioxide (CC^), methane (CH4), nitrous
oxide (N2O) and hydrofluorocarbons (HFCs) were quantified. In addition to reducing the
emissions of greenhouse gases, today's rule would also influence the emissions of "criteria"
air pollutants, including carbon monoxide (CO), fine particulate matter (PM^s) and sulfur
dioxide (SOx) and the ozone precursors hydrocarbons (VOC) and oxides of nitrogen (NOx);
and several air toxics (including benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and
acrolein).
Most analyses in this chapter of the RIA were updated between proposal and final
rulemaking. Most significantly, as a result of public comments and updated economic data,
the attribute based CC>2 curves have been revised, as discussed in detail in Section II.B of this
preamble and Chapter 2 of the Joint TSD. This update in turn affects costs, benefits, and
other impacts of the final standards. Thus EPA's overall projection of the impacts of the final
rule standards have been updated and the results are different than for the NPRM, though in
general not by a large degree.
Beyond updated CO2 curves, other new inputs includes revised sales projections of the
MY 2012-2016 fleet, and updated economic input data. All changes to inputs are documented
in the TSD, and are further described in this document.
Downstream (tailpipe) emission impacts were developed using 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). Non-CO2
5-1
-------
Regulatory Impact Analysis
emissions were calculated using data from EPA's Motor Vehicle Emission Simulator
(MOVES2010).
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.6 Based on these analyses, the
control programs set forth in this chapter would account for 307 MMT COiEQ of annual
GHG reduction in the year 2030 and 506 MMT per year in 2050. Fuel savings resulting from
the GHG standards are projected at 41.5 billion gallons of fuel savings in Calendar Year 2050
(Table 5-1).
Table 5-1 Impacts of Program on GHG Emissions and Fuel Savings
CALENDAR
YEAR
2020
2030
2040
2050
ANNUAL GHG
REDUCTION (CO2
EQ MMT)
156.3
307.4
401.5
505.9
FUEL SAVINGS
(MILLION BARRELS
PER DAY OF
GASOLINE
EQUIVALENT)
0.8
1.6
2.1
2.7
ANNUAL FUEL
SAVINGS
(BILLION GALLONS
OF GASOLINE
EQUIVALENT)
12.6
24.7
32.6
41.5
The emissions of non-GHG air pollutants due to light duty vehicles are also expected
to be affected by today's final rule. These effects are due to changes in driver behavior (the
"rebound effect")A and also reflect ethanol volume assumptions that are not due to the new
GHG vehicle standards. The delta values shown here include both upstream and downstream
contributions.
A A rebound effect of 10% is used in this analysis. See section 5.3.3.1.1 for a brief definition of rebound, and
chapter IV of the joint Technical Support Document for a more complete discussion.
5-2
-------
Emissions Impacts
Table 5-2 Impacts of 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 Particulate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR
YEAR
2020
-95.1
760.0
0.8
-889.9
3,980.3
-49.4
-5,916.1
-2,402.9
-13,853.4
-60,305.4
CHANGE
VS. 2020
REFEREN
CE
-0.38%
2.26%
0.01%
-0.48%
0.01%
-0.06%
-0.02%
-0.03%
-0.42%
-0.51%
CALEND
AR YEAR
2030
-21.1
668.1
4.7
-523.1
170,648.6
15.1
-21,845.0
-4,574.8
-27,492.8
-115,816.5
CHANGE
VS. 2030
REFEREN
CE
-0.10%
2.18%
0.07%
-0.29%
0.56%
0.02%
-0.07%
-0.05%
-0.82%
-1.02%
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 final rule. These results, including both
upstream and downstream GHG contributions, are presented below (Table 5-3).
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)
88.8
130.2
174.2
244.2
324.7
962.0
Lifetime Fuel Savings
(Billion Gallons Of
Gasoline Equivalent)
7.3
10.5
13.9
19.5
26.5
77.6
Lifetime Fuel Savings
(Million Barrels of
Gasoline Equivalent)
173.1
250.35
330.5
464.7
630.7
1,849.3
5-3
-------
Regulatory Impact Analysis
5.2 Introduction
5.2.1 Scope of Analysis
Today's program finalizes new standards for the greenhouse gas (GHG) emissions of
light duty vehicles from model year 2012 through model year 2016. The 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 final rule 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 final rule.
This chapter describes the development of inventories for emissions of the gaseous
pollutants impacted by the 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 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 final rule on greenhouse gases in calendar years 2020, 2030, 2040 and
2050. Non-greenhouse gas inventories are shown in 2020 and 2030. The program was
quantified as the difference in mass emissions between the standards and a reference case as
described in Section 5.3.2.2.
Table 5-4 Processes Considered
PROCESS
Crude Oil Extraction
Crude Oil Transport
Oil Refining
Fuel Transport and Distribution
Fuel Tailpipe Emissions
Air Conditioning System Leakage
UPSTREAM / DOWNSTREAM
Upstream
Upstream
Upstream
Upstream
Downstream
Downstream
Inventories for the four greenhouse gases carbon dioxide (CO2), methane (CFLO, 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. Inventories for the non-GHG pollutants 1,3-butadiene, acetaldehyde, acrolein,
benzene, carbon monoxide (CO), formaldehyde, oxides of nitrogen (NOx), particulate 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 final rule is new standards for
tailpipe emissions produced during vehicle operation. Absolute reductions from tailpipe
GHG standards are projected to grow over time as the fleet turns over to vehicles affected by
5-4
-------
Emissions Impacts
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 COi emitting vehicles.
As described herein, the downstream reductions in greenhouse gases due to the 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 HFC s (direct improvement) and reducing fuel consumption
by increasing the efficiency of the air conditioning system (indirect).
Due to the rebound effect,8 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. The implications for non-GHG emissions of
changes in fuel supply were analyzed for the final rulemaking and are discussed in Section
5.3.3.5.
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. 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 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.7
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).8
The global warming potentials (GWP) used in this analysis are consistent with the 2007
Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4). At
this time, the 1996 IPCC Second Assessment Report (SAR) global warming potential values
1 Described in Joint TSD Chapter 4
5-5
-------
Regulatory Impact Analysis
have been agreed upon as the official U.S. framework for addressing climate change and are
used in the official U.S. greenhouse gas inventory submission to the United Nations climate
change framework. This is consistent with the use of the SAR global warming potential
values in current international agreements.
Table 5-5 Global Warming Potentials for the Inventory GHGs
Gas
CO2
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) in
conjunction with EPA's Motor Vehicle Emission Simulator (MOVES2010). Broadly
speaking, OMEGA is used to predict the most likely paths by which manufacturers would
meet tailpipe COi 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 impacts (emission reductions,
monetized co-benefits) of the analyzed program.
The OMEGA post-processor was updated with emission rates MOVES2010.9'10 COi
emission and fuel consumption rates are drawn from OMEGA results, with all co-pollutant
emission rates derived from the MOVES2010 emission rate database. Air conditioning
inventories (including HFC and COi contributions) were separately calculated in spreadsheet
analyses, and are based on previous EPA research.11 Both MOVES and OMEGA are
published, publicly available models and continue to be actively developed.12'13 No public
comments were received on the selection of either MOVES or the OMEGA post-processor for
calculating the impacts of this rule.
Upstream emissions were calculated using the same tools as were used for the
Renewable Fuel Standard 2 (RFS2) rule analysis,14 but for the current analysis it was assumed
that all impacts are related to changes in volume of gasoline produced and consumed, with no
5-6
-------
Emissions Impacts
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.15'16 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 rulemaking 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. No public comments were received on EPA's use of the modified version of GREET in
this analysis.
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 analysis, and each is
described below.17 The two scenarios shown are differentiated by their regulatory COi
emission standards. The reference scenario CC>2 emissions are based upon the National
Highway Traffic Safety Administration (NHTSA) Model Year 2011 Corporate Average Fuel
Economy (CAFE) standards,18 while the control scenario CC>2 emissions are based upon the
program set forth 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.
As in the proposal, for this analysis we attribute decreased fuel consumption from this
program to gasoline only, while assuming no effect on volumes of ethanol and other
renewable fuels because they are mandated under the Renewable Fuel Standard (RFS2).
However, because this rule does not assume RFS2 volumes of ethanol in the baseline, the
result is a greater projected market share of E10 in the control case.19 In fact, the GHG
standards will not be affecting the market share of E10, because EPA's analysis for the RFS2
rule predicts 100% E10 penetration by 2014.20
In the proposal, EPA stated these same fuel assumptions and qualitatively noted that
there were likely unquantified impacts on non-GHG emissions between the two cases. In RIA
Chapter 5, EPA indicated its plans to quantify these impacts in the air quality modeling and in
the final rule inventories. Upstream emission impacts depend only on fuel volumes, so the
impacts presented here reflect only the reduced gasoline consumption.
The inventories presented in this rulemaking include an analysis of these fuel effects
which was conducted using EPA's Motor Vehicle Emission Simulator (MOVES2010). The
most notable impact, although still relatively slight, is a 2.2 percent increase in 2030 in
national acetaldehyde emissions over the baseline scenario. It should be noted that these
emission impacts are not due to the new GHG vehicle standards. These impacts are instead a
consequence of the assumed ethanol volumes. This program does not mandate an increase in
5-7
-------
Regulatory Impact Analysis
E10, nor any particular fuel blend. The emission impact of this shift was also modeled in the
RFS2 rule.
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. 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 less than 1% (Section 5.3.3.5).
The relationship between fuel composition and emission impacts used in
MOVES2010 and applied in this analysis match those developed for the recent Renewable
Fuels Standard (RFS2) requirement, and are extensively documented in the RFS2 RIA and
supporting documents.21
5.3.2.1 Sales and Fleet Composition
Fleet composition has a significant effect upon the impacts of the program.
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.22 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 201 1)
EPA Program
(MY 2012+)
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
As explained in section II.B of the preamble to the final rule and chapter 1 of the Joint
TSD, EPA updated its fleet projection for the final rulemaking analysis.23 As a result of this
change, all calculations which depend upon fleet composition (ie, emission inventories,
impacts of flexibilities, and fuel savings) were updated from those in chapter 5 of the RIA.
Total volumes of projected sales of classic cars and trucks for calendar years 2012-
2035 were drawn from the Energy Information Administration Annual Energy Outlook
(AEO) 2010 Early Release projection (December 2010).24 The AEO 2010 Early Release is an
5-8
-------
Emissions Impacts
update of the April 2009 AEO projections used in the proposal analysis.25 Based on EPA
analysis of the projected MY 2012-2016 fleet,26 approximately 20% of the classic truck fleet
is anticipated to be reclassified as Passenger Automobiles under the new standards. The
AEO 2010 sales projections, which are based on the classic fleet, were then reclassified using
PA and NPA definitions by shifting 19.91% of AEO's truck sales projection to car sales. For
calendar years 2035-2050, which are beyond the scope of AEO's projections, 0.88% annual
growth in the sales of cars and trucks was assumed. The annual growth rate of 0.88% is the
average year-on-year sales growth projected by AEO from years 2017-2035.
A more complete discussion of the process for developing the MY 2012-2016 fleet is
available in TSD chapter 1.
Table 5-7 Projected Total Vehicle Sales and Car Fractions
Total Light
Duty Sales
Classic Car
Fraction
PA Fraction
PAs Sold
Model Year
2012
14,921,031
51.8%
59.2%
7,922,992
Model Year
2013
15,835,190
52.9%
61.1%
9,123,197
Model Year
2014
16,178,725
54.3%
61.9%
9,797,738
Model Year
2015
16,452,676
55.8%
63.2%
10,231,974
Model Year
2016
16,501,102
57.1%
64.6%
10,627,055
5.3.2.2 Fleet Average CO2 Targets
In this section, the term "target" is used to refer to the output of the footprint equations
described in Preamble Section II. The term "achieved emission level" is similar, but includes
the impacts of program flexibilities.
As documented in Preamble Section II, under both reference and control scenarios,
each manufacturer has a unique fleet average target based on their vehicle footprints and
production.
Fleet average targets are calculated by weighting the individual PA and NPA targets
by the respective proportions of anticipated production (Section 5.3.2.1). These COi emission
values are unadjusted values (i.e. in CAFE space), so they are lower than the anticipated on-
road emissions. In all scenarios, post- 2016 vehicles are assumed to maintain model year
2016 emissions. Because the fleet composition continues to change post-MY 2016, the fleet
average emission level continues to vary. No public comments were received on this
methodology.
Below, PA and NPA tailpipe COi fleet average emission targets and achieved
emission levels during MY 2012-2016 are shown for reference and control scenarios.
5-9
-------
Regulatory Impact Analysis
5.3.2.2.1 Reference Case
5.3.2.2.1.1 CO2 Emission Targets
The reference scenario targets 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 joint TSD and chapter 4 of the RIA).27 Average car and truck fuel
economy targets were calculated from the coefficients in the MY 2011 rule and the projected
MY 2012-2016 fleet.28'29 Average fuel economy targets were calculated for each
manufacturer's fleet, and then combined based on projected sales.
A ratio of 8,887 grams of COi emitted per gallon of gasoline was used to convert to
the calculated fuel economy standards to CC>2 (gram/mile) emission factors. The basic
derivation of the 8,887 factor can be seen in previous EPA publications.
30
Minor changes in the emission targets are due to projected changes in the average new
vehicle footprint between 2012 and 2016 (Table 5-8).
Table 5-8 Reference Case Average Emission Targets (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
2015
2016
PA
EMISSION
LEVEL
292
291
291
292
292
NPA EMISSION
LEVEL
365
365
364
364
364
MY EMISSION
LEVEL
320
319
318
317
316
5.3.2.2.1.2 Achieved CO2 Emission Levels
The emission targets shown in Table 5-8 do not reflect the impact of several program
flexibilities in CAFE program, 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 otherwise allows.
Additionally, some manufacturers overcomply with the standards, and this overcompliance is
not reflected in the CAFE targets.
While the CAFE program is complex, the most significant portions of the program
flexibilities were accounted for. In this analysis, manufacturer overcompliance, credit trading,
FFV credits, and fine paying manufacturers were included. Credit banking was excluded.
In general, achieved emission levels were estimated by beginning with the more
stringent of either (A) a manufacturer's CAFE target (in COi space) or (B) estimated actual
5-10
-------
Emissions Impacts
MY 2008 COi emission levels 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 it is
anticipated that they will take advantage. Consistent with the use of the MY 2011 standards,
the credits and trading limits available for MY 2011 were 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 Transfers
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 target (in COi 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 "transfered" 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 transfered 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 transfer 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 achieved levels based on program flexibilities and manufacturer
Overcompliance are shown in
Table 5-10.
5-11
-------
Regulatory Impact Analysis
Table 5-9 Impacts of credits (grams/mile CO2 EQ)
MODEL
YEAR
2012
2013
2014
2015
2016
OVERCOMPLIANCE,
CREDITS AND
TRANSFERS
-5.1
-6.1
-6.4
-6.7
-7.0
FFV
7.5
7.1
6.6
6.4
6.3
FINES
1.0
0.6
0.2
0.1
0.1
NET
3.5
1.5
0.3
-0.2
-0.6
Table 5-10 Reference Case Achieved Emissions (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
286
284
283
283
283
ANTICIPATED
NPA EMISSION
LEVEL
383
381
379
378
378
ANTICIPATED
MY EMISSION
LEVEL
324
321
318
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 establishing a footprint attribute based
function in order to determine the CC>2 (gram/mile) emission standard for a given vehicle.
The piecewise linear function used by EPA is documented in Section II.B of the preamble to
the final rule. 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 targets for the
MY2012-2016 vehicles (Table 5-11)."
31
Table 5-11 Control Case Average Emission Targets (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
PA
EMISSION
LEVEL
263
256
247
NPA EMISSION
LEVEL
346
337
327
PROJECTED
MY EMISSION
TARGET
295
286
276
5-12
-------
Emissions Impacts
2015
2016
236
225
312
298
263
250
5.3.2.2.2.2 Achieved CO2 Emission Levels
Just as with the reference scenario, the control case emission targets (Table 5-11) do
not include the effect of several flexibilities built into the EPA program.
The same basic methodology was used to calculate achieved fleet emission levels for
the control case as in the reference case. In general, achieved emission levels were estimated
by beginning with the more stringent of either (A) a manufacturer's calculated footprint-based
emission target or (B) the estimated achieved COi level based on the EPA fleet data file.
Using that starting point, each manufacturer's emissions were increased by the impact of the
credits which we anticipate manufacturers will utilize. Manufacturers were always assumed to
perform at least as well as they did in 2008.
Overcompliance and Credit Transfers
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 their target.
The achieved Overcompliance on either PA or NPA could be "transfered" 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 program, there are no limits within the light duty fleet on such trading.c This
transferance 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 transfers are VMT weighted, the resulting changes will be
essentially environmentally neutral on both GHG and criteria pollutants.
Flex Fueled Vehicles
The flex fueled vehicle credit,consistent with the final rule is set at 1.2 MPG for MY
2012-2014, 1.0 MPG for MY 2015, and 0 MPG for MY 2016+. See also preamble section
c Preamble section III.B and III.C discusses credit transfers in more detail, including limits on credit life and
various other restrictions.
5-13
-------
Regulatory Impact Analysis
III.C.2 As in the reference case, it was assumed that the manufacturers General Motors,
Chrysler and Ford would utilize 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 COi per mile for the fleet, while direct A/C
credits were set at 6.3 grams COi per mile for PA and 7.8 grams COi per mile for NPA). In
the proposal, we noted the inconsistent values between the direct A/C credit presented here,
and the direct A/C credit discussed in RIA Chapter 2. We corrected this minor inconsistency
for this FRM analysis. EPA assumed market penetration of the technology according to Table
5-17. A more complete discussion of the A/C credit program and inventories is provided in
section 5.3.3.2, as well as RIA chapter 2.
Temporary Lead Time Allowance Alternative Standards (TLAAS)
In response to public comment, the TLAAS program includes certain additional
features which expand its range. See preamble section III.B.5 and Appendix A to this RIA
chapter. Specifically, the TLAAS program has been expanded into two distinct tiers, which
are manufacturers with fewer than 400,000 sales and manufacturers with fewer than 50,000
sales. Manufacturers with less than 5,000 vehicles in sales were also temporarily exempted
from this rulemaking. A brief summaryof the inputs used in this analysis appear below. For
more on the TLAAS program, please see Appendix A to this RIA chapter.
For the larger manufacturers, we assumed that every potentially eligible manufacturer
utilized the TLAAS program. Each qualifying manufacturer was assumed to use the full
vehicle allocation according to the default production schedule shown in Section III.B.5 of the
proposal preamble and reproduced in Table 5-12.
Table 5-12 - TLAAS default production schedule
MODEL
YEAR
Sales Volume
2012
40,000
2013
30,000
2014
20,000
2015
10,000
5-14
-------
Emissions Impacts
The allocation was split evenly between cars and trucks for each manufacturer. For
these companies, this vehicle allocation was assumed to emit as much CO2 per mile as the
highest emitting car or truck in each manufacturer's fleet.
For the smaller manufacturers, the program was expanded to allow 50,000 vehicles in
2016, and 200,000 vehicles though 2015. These fleets were assumed to gradually phase into
compliance. These TLAAS fleets are assumed to emit 1.25x more emissions than the
manufacturer's sales weighted target in 2012, and by 2016, they were assumed to emit 1.05x
more emissions.
In each case, the TLAAS vehicles were then proportionally averaged into the
manufacturer's achieved emission level.
5-15
-------
Regulatory Impact Analysis
The aggregate impacts of these program flexibilities are listed in Table 5-13.
Table 5-13 Estimated Impacts of Program Flexibilities (grams/mile CO2 EQ)
MODEL
YEAR
2012
2013
2014
2015
2016
OVERCOMPLIANCE,
CREDITS AND
TRANSFERS
-0.1
0.0
0.0
0.0
0.0
FFV
6.5
5.8
5.0
3.7
0.0
DIRECT
A/C
1.7
2.7
4.1
5.5
5.8
INDIRECT
A/C
1.4
2.3
3.4
4.6
4.8
TLAAS
1.2
0.9
0.6
0.3
0.1
NET
10.7
11.7
13.1
14.0
10.7
Based on these impacts, the achieved emission level by PA, NPA and fleet are
displayed in Table 5-14. 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 5-14 Federal GHG Program Anticipated Emission Levels (grams/mile CO2)
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
270
264
258
248
236
ANTICIPATED
NPA EMISSION
LEVEL
365
354
344
330
309
ANTICIPATED
MY EMISSION
LEVEL
307
298
290
277
261
Table 5-14 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
here. However, because the trading is VMT weighted, it is environmentally neutral and has
no GHG emissions impacts.
As in the proposal, the OMEGA also predicts slight undercompliance in 2016 for
several manufacturers, while the results presented here assume full compliance. The net
undercompliance is approximately 0.8 grams in 2016. A more complete discussion of the
OMEGA cost modeling is available in RIA chapter 4.
5.3.3 Calculation of Downstream Emissions
As stated in Section 5.1, the downstream analysis conducted in the proposal has been
updated in the analysis shown here. To reiterate, the 2012-2016 COi standards (i.e., the
5-16
-------
Emissions Impacts
attribute-based curves for cars and light trucks) were revised slightly,0 and several tools were
updated; Draft MOVES2009 was updated to MOVES2010 and minor changes were made to
the OMEGA post-processor. Beyond these changes, the analysis of GHG emissions was
similarly conducted between NPRM and FRM. While public comments were received on
several of the economic inputs used in the modeling (see TSD chapter 4), no substantive
comments were received concerning the methodology or resulting inventories.
As mentioned in Section 5.2.2, the analysis of non-GHG emissions was updated to
include fuel effects. The FRM upstream analysis was updated with the new fuel savings
volumes, but is otherwise unchanged.
A model year lifetime analysis, considering only the five model years
specifically regulated by 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 Calculation of Tailpipe COi Emissions
The fleet inputs (achieved COi emission levels by model year and vehicle
sales) were incorporated into a spreadsheet along with emission rates derived from
MOVES2010 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.
Two basic elements feed into OMEGA's calculation of vehicle tailpipe
emissions. These elements are VMT and emission rates.
Total Emissions = VMT ^les * Emission rate grams/mile
Equation 1 - 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.
D See Preamble Section II.B
5-17
-------
Regulatory Impact Analysis
5.3.3.1.1 BaseVMT
The downstream analysis is based upon a "bottom-up" estimate of total VMT
and vehicle population. The VMT inputs are documented more fully in 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
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 accumulation schedule for that vehicle
class, model year, and age.
Future vehicle sales were drawn from AEO 2010 Early Release (as discussed
in Section 5.3.2.1), while historic vehicle sales are drawn from the Transportation
Energy Data Book,32 Post MY 2011 vehicles were reclassified in order to correspond
to the PA/NPA definitions.
As described in the TSD, mileage accumulation by age was calculated using
inputs from the NHTSA "Vehicle Survivability and Travel Mileage Schedules" and
additional inputs unique to this analysis.33'34 In brief, a 1.15% per vehicle annual
VMT growth rate was assumed, but additional factors such as achieved fuel
consumption and the price of gasoline also contributed to the precise schedule for each
MY.
The vehicle 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 5-15).35
5-18
-------
Emissions Impacts
Table 5-15 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
5-19
-------
Regulatory Impact Analysis
A complete discussion of the derivation of the MY specific VMT schedules is
provided in joint TSD chapter 4.
5.3.3.1.2 Rebound
The tailpipe CC>2 standards are expected to result in greater fuel efficiency. Per
the discussion of the rebound effect in the 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:
A%VMTreb = -REB *
(FleetFCM-FleetFCnJ
FleetFCold
Equation 2 - 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 supporting that figure
is explored in greater depth in chapter 4 of the 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, CO2 emission rates are derived from the achieved
vehicle emission levels in
Table 5-10 & Table 5-14, SOi emission rates are derived from fuel sulfur levels, and
the emission rates for the remaining pollutants are derived from the MOVES2010 database.
For a more complete discussion of these emission rates, please refer to joint TSD chapter 4.36
EPA is not projecting any reductions in tailpipe CH4 or N2O emissions as a result of
these emission caps.Similar to other pollutants, there are downstream emission impacts due to
changes in fuel supply and increased driving (rebound), as well as upstream impacts due to
decreases in fuel production, transport, and distribution.
5-20
-------
Emissions Impacts
5.3.3.1.4 Tailpipe CO2 Emissions from Vehicles
CO2 emission rates were derived from the achieved COi emission levels in
Table 5-10 & Table 5-14. Previous EPA analysis has shown that an
approximately 20% gap exists between CAFE space fuel economy and on-road fuel
economy.37 The on-road gap is more fully documented in the joint TSD chapter 4.
The 20% gap, while approximate, includes average effects of energy
consumption 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 COi
emission impacts of running a mobile vehicle air conditioning system. In this
analysis, COi emissions are assumed to remain unchanged throughout the vehicle's
lifetime.
By dividing a CAFE-space CC>2 emission rate by one minus the on-road gap,
one can approximate the actual on-road CCh emission rate experienced by drivers, and
this analysis used this means of reflecting the on-road gap. By including VMT, we
estimate the on-road tailpipe COi emissions.
On road tailpipe CO2 emissions =
Achieved COi Emission Level / (1-on-road gap) x VMT including rebound
Equation 3 - Tailpipe CO2 Emissions Excluding A/C
Based on Equation 3, the baseline CO2 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 5-16 Tailpipe CO2 Emissions including Baseline A/C Usage (MMT)
Tailpipe COi Emissions
(Reference)
A Tailpipe COi Emissions
(Control) including 10%
rebound
A Tailpipe COi Emissions
due to 10% rebound
2020
1,173
-101.2
10.3
2030
1,313
-199.6
19.9
2040
1,609
-263.4
26.1
2050
2,030
-335.3
33.2
5.3.3.2 Air Conditioning Emissions
Outside of the tailpipe COi 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 COi. With regard to air conditioning-related emissions, significant
5-21
-------
Regulatory Impact Analysis
opportunity exists to reduce HFC emissions from refrigerant leakage (direct emissions) and
i from A/C induced engine loads (indirect emissions).
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 RIA Chapter 2.
EPA's analysis, shown in RIA chapter 2, indicates that A/C- related emissions
accounted for approximately 8% of the GHG emissions from in-use cars and light trucks in
2005. EPA is finalizing credit provisions which we expect all manufacturers to utilize which
are expected to reduce direct leakage emissions by 50% and to reduce indirect A/C emissions
(A/C related CCh tailpipe 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 CC^in that they would be achieved through specific hardware and controls.
The anticipated phase-in of air conditioning controls is shown in Table 5-17. The 85%
cap is roughly linearized across the five year period (Table 5-17). Because HFC leakage is
somewhat independent of vehicle miles traveled, the HFC fraction is based upon the
proportion of new vehicles that have HFC leakage containment technology. By contrast, the
indirect A/C reduction fraction is dependent upon the travel fraction, and is proportional to the
VMT traveled by vehicles with the control technology.
5-22
-------
Emissions Impacts
Table 5-17 - AC Control by Model Year (Reduction from Base Emissions)
Market Penetration of technology
HFC Reduction %
Indirect Reduction %
MY
2012
25 %b
-13%
-10%
MY
2013
40%
-21%
-16%
MY
2014
60%
-30%
-24%
MY
2015
80%
-40%
-32%
MY
2016+
85%
-43%
-34%
5.3.3.2.1 Direct A/C (HFC) Emissions
38
The projected HFC baseline inventories are derived from previous EPA analyses.
The methodology used in the proposal was updated with the new estimates of vehicle sales
and miles traveled.
As noted, HFC emissions are a leakage type emission, similar to other evaporative
emissions from a vehicle.39 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)40 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. As noted in the proposal and reiterated here, preliminary EPA analysis indicates
that air conditioner charge size is decreasing, which implies that the analysis presented here
may 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 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 4).
B In Preamble Section III, the expected penetration of A/C control technology is shown to be 28% in MY 2012.
The slightly lower penetration number used in the emission modeling indicates a slight underestimation of the
emission reductions from MY 2012, and consequently the benefits from this rule.
5-23
-------
Regulatory Impact Analysis
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 4 - HFC Inventory Calculation
Table 5-18 shows baseline HFC inventory and control scenario reductions.
Table 5-18 HFC (Direct A/C) Emissions
Calendar
Year
2010
2020
2030
2040
2050
Baseline HFC
(MMT CO2EQ)
56.9
61.3
67.8
73.7
80.4
Reduction From
Baseline
(%)
0%
-22%
-38%
-42%
-42%
Reduction from
Baseline
(MMT CO2EQ)
0.0
-13.3
-26.0
-30.9
-34.2
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 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 RIA 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 put forth 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 methodology used in the proposal was updated with the new estimates of vehicle
sales and mileage traveled. The OMEGA post processor was used to calculate the
contribution of the indirect A/C program to the overall inventory. Reference and control
scenario emissions attributable to indirect A/C systems are shown in Table 5-19.
Table 5-19 -Indirect A/C Emissions
Calendar
Year
2010
2020
2030
2040
Baseline Indirect A/C
(MMT CO2EQ)
53.1
53.6
63.1
78.5
Reduction From
Baseline
(%)
-0%
-20%
-32%
-34%
Reduction from
Baseline
(MMT CO2EQ)
0
-10.6
-20.2
-26.5
5-24
-------
Emissions Impacts
2050
99.3
-34%
-33.8
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 Methane (CH4) and Nitrous Oxide (NiO) Emissions from Vehicles
MOVES2010 does not include fuel effects for either nitrous oxide or methane emissions.
Therefore, the only modeled difference in NiO and CH4 between control and reference cases
are emissions which occur during rebound driving. These emissions, like all rebound
emissions, were calculated in the modified OMEGA post-processor.
The reference inventories shown in Table 5-20 were calculate using MOVES2010 as
described in Section 5.3.3.5.1.
Table 5-20 Downstream CH4and N2O Emissions (Metric Tons)
Gasoline Vehicles
CH4
N2O
Diesel Vehicles
CH4
N2O
Reference Emissions
2020
41,828
29,898
661.6
829.8
2030
44,464
21,620
810.3
999.3
Control Emissions
(including rebound)
2020
42,130
29,898
661.7
830.0
2030
45,096
21,904
810.6
999.7
Delta Emissions
2020
302
135
0.1
0.2
2030
632
284
0.3
0.4
5.3.3.4 Fuel Savings
The 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 A/C
use) using a conversion factor of 8887 grams of CO2 per gallon of gasoline. All fuel saved is
considered 100% gasoline without any oxygenate'F'41
Fuel savings were calculated from total tailpipe COi avoided (including COi due to
driving and indirect A/C) using a conversion factor of 8887 grams of CC>2 per gallon of
gasoline. 42
F Based on the documentation of the on-road gap, it would be justifiable to assume an ethanol percentage of
approximately 2.3%. This volume of ethanol would result in a total energy difference of less than 1%. See the
fuel labeling rule technical support document, EPA420-R-06-017, for further details.
5-25
-------
Regulatory Impact Analysis
Table 5-21 - 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
137.9
-12.6
1.2
-1.2
2030
154.8
-24.7
2.2
-2.3
2040
190
-32.6
2.9
-3.0
2050
239.6
-41.6
3.7
-3.8
5.3.3.5 Downstream Criteria and Air Toxic Emissions
This rule affects tailpipe co-pollutant emissions in two significant ways. The first,
modeled in the proposal, is an increase in emissions due to the rebound effect. The second
effect, newly analyzed here, is an increase in the market share of ethanol blended gasoline.
As modeled, this rule will reduce the consumption and production of gasoline (EO), while the
production of ethanol is held constant due to the Renewable Fuel Standards. Consequently,
the fraction of fuel which is blended to 10% ethanol (E10) is assumed to increase. However,
the increased E10 market share is projected to occur regardless of this rule; in the RFS2
analysis we project 100% E10 by 2014. These fuel effects were not quantified in the proposal;
as a result the proposal emission inventories differ from those shown here.
For today's analysis, MOVES2010 was used to generate base inventories for both
reference and control fuel supplies using a single base VMT. Using the control scenario fuel
supply, the OMEGA post-processor provided rebound emission quantities.
5.3.3.5.1 Base Criteria and Air Toxic Emission Inventories
MOVES2010 was run in the following manner in order to provide base inventories for
both reference and control cases.
The fuel supplies in each case were calculated by sequentially:
A) estimating the light duty energy demand in 2020 and 2030 based on the
energy consumption projections from this rule.
B) determining the total energy demand from light duty, heavy duty, motorcycle
and non-road sources from AEO and other reference sources
C) determining the total ethanol volume from AEO 2007, which does not include
increased renewable fuel volumes due to EISA.0
0 Due to the long lead times required for this analysis, it was completed before the second Renewable Fuel
Standard (RFS2) was signed. The increased renewable fuel volume attributable to this regulation is therefore not
assumed in this analysis.
5-26
-------
Emissions Impacts
D) calculating E10 market share
The resulting 2020 and 2030 E10 market shares are shown in Table 5-22.
Table 5-22 E10 market shares
Calendar
Year
2020
2030
Reference Case
E10 fraction
85%
82%
Control Case
E10 Fraction
92%
97%
Delta
7%
15%H
For simplicity, we used the same MOVES fuel supply table as used in the Renewable
Fuel Standard 2 analysis as our reference case fuel supply. This table is approximately 90%
E10. For the control case, we created a table where 100% of the fuel supply was E10. These
tables were used in both of the modeled years (2020 and 2030). This slightly overstates the
delta in E10 usage in 2020 and slightly understates the delta in 2030. The simplified analysis
still captures the approximate deltas shown in Table 5-22.
To maintain consistency with the MOVES runs conducted for the Air Quality analysis,
temperatures from 2005 were also input to MOVES, and used in both years of analysis.
Four separate MOVES runs were conducted. All valid sourcetype/fuel type
combinations in the MOVES2010 database were included in the MOVES runs. For each of
the control case and reference case, one run was used to calculate evaporative emissions, and
one run was set for all other processes (running exhaust, start exhaust, brakewear, tirewear,
crankcase, refueling, and extended idle). Diesel toxic emissions were not produced by the
MOVES model, but were post-processed from VOC emissions using published ratios
As the production of petroleum based fuels decreases, the market share of E10 is projected to
gradually increase. E10 has slightly less energy than EO, a consequence of which may be a slight
reduction in the quantity of retail gasoline gallons saved by this rule. The total energy savings would
remain as predicted by this rule.
Assuming that a gallon of ethanol contains approximately 77,000 BTU of energy and that a gallon of
gasoline contains approximately 115,000, a gallon of E10 contains 3.3% less energy than a gallon of
gasoline. A 15% increase in E10 market share in 2030, as described in table 5-22, would indicate that
the average gallon of gasoline sold in the control case in 2030 would contain 0.5% less energy than in
the reference case. This difference in energy content would be less in the near term, before the
program is fully phased in.
All else being equal, the difference in energy content would result in additional gallons of fuel being
purchased to meet the energy demands of the control case. Assuming that gasoline prices would not
be affected by ethanol or energy content, this would result in a very slight overestimate of the
monetized fuel savings predicted by this rule and discussed in RIA Chapter 6.
5-27
-------
Regulatory Impact Analysis
identical to those in MOVES.43 All emission results except for evaporative emissions were
post-processed and scaled based on the ratios of VMT between the MOVES output and the
VMT used for air quality modeling. Evaporative emissions, which are largely dependent on
vehicle population, were not scaled. As the air quality modeling was only conducted in 2030,
2020 VMT was created by scaling the air quality modeling VMT by the ratio of 2020 to 2030
VMT in MOVES. The resulting factor (82%) was universally applied to the 2030 VMT to
produce 2020 VMT for each sourcetype.
The VMT developed for air quality modeling is more completely documented in
Section 5.8.
5.3.3.5.2 Criteria and Air Toxic Emissions due to the Rebound Effect
As a result of the additional rebound VMT, the downstream emissions of several co-
pollutants increase in the control case. The emissions due to rebound were calculated in the
OMEGA post-processor in a similar manner to the CO2 emissions. Rebound VMT was
broken into distribution by vehicle age and was then multiplied by the appropriate emission
factor. These emissions by age were then summed by calendar year (Equation 5).
Emissionscaiendar Year = Xcaiendar Year (Rebound VMT by Age * Emission Factor by Age)
Equation 5 - Emissions by Calendar Year
The EPA reference fleet assumes a small number of diesel vehicles are sold in each
year (approximately 20 thousand vehicles out of approximately 13-16 million). For the
analysis of criteria emissions due to the rebound effect, it was assumed that 0.5% of new light
duty vehicles sold were diesels. Because diesel fueled vehicles are subject to the same Tier 2
emission standards as gasoline fueled vehicles, the emission rates of criteria pollutants are
similar.1
5.3.3.5.3 Tabulation of Downstream Criteria and Air Toxic Impacts
This section contains a table of the downstream criteria and air toxic emissions.
1 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.
5-28
-------
Emissions Impacts
Table 5-23 - Downstream non-GHG Emissions (Short Tons)
Gasoline Vehicles
1,3-Butadiene
Acetaldehyde
Acrolein
Benzene
CO
Formaldehyde
NOX
PM2.5
S02
VOC
Diesel Vehicles
1,3-Butadiene
Acetaldehyde
Acrolein
Benzene
CO
Formaldehyde
NOX
PM2.5
SO2
VOC
Reference Emissions
CY 2020
4,777
10,805
569
30,633
20,764,531
11,268
1,493,306
40,685
22,130
1,290,008
1,397
5,562
741
2,620
612,037
15,279
1,502,844
41,483
4,565
203,977
CY 2030
3,448
7,909
419
22,048
20,615,741
8,196
1,059,567
42,855
25,700
1,006,387
1,434
5,709
761
2,689
567,933
15,684
1,192,334
15,678
5,538
209,384
Control Emissions
(including rebound)
CY 2020
4,683
11,571
571
29,882
20,774,455
11,270
1,506,643
40,915
20,103
1,294,309
1,397
5,562
741
2,620
612,270
15,280
1,502,917
41,484
4,565
203,994
CY 2030
3,429
8,590
426
21,799
20,797,866
8,311
1,075,547
43,455
21,451
1,017,550
1,434
5,709
761
2,690
568,595
15,686
1,192,498
15,680
5,538
209,428
Delta Emissions
CY 2020
-93.7
766.6
1.7
-750.4
9,924.0
1.4
13,336.7
229.8
-2,027.7
4,301.6
0.2
0.2
0.1
0.3
232.2
0.7
73.7
1.2
CY 2030
-18.5
681.0
6.4
-249.2
182,125.6
114.6
15,979.4
599.7
-4,248.5
11,163.3
0.4
0.5
0.2
0.9
662.2
1.7
163.2
2.6
Attributed to gasoline
17.1
44.1
In summary, the downstream emissions of the criteria pollutants CO, NOx, PM2.5, and
VOC increase due to the additional rebound VMT. SO2 emissions decrease because the CO2
standards lead to a decreased volume of fuel consumption and less resulting emissions of
sulfur compounds.
Air toxic emissions, which are sensitive to fuel effects, vary more between cases.
Acetaldehyde emissions increase roughly proportionally to the increase in ethanol penetration.
Similarly, benzene and 1,3 butadiene decrease proportionally to the decrease in gasoline
emissions. These changes are the result of our ethanol volume assumptions and are not due to
the new GHG vehicle standards. For a more complete discussions of ethanol effects on air
toxic emissions, please refer to the EPA RFS2 analysis.
44
As will be 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 those pollutants
5-29
-------
Regulatory Impact Analysis
such as carbon monoxide (CO), acetaldehyde, and formaldehyde, where a relatively small of
US emissions comes from upstream sources.
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."45 The results of this analysis are shown in Table 5-30. No public
comments were received on the methodologies used in the calculation of upstream
inventories.
5.4 Greenhouse Gas Emission Inventory
This section presents total program calendar year impacts by sector (Table 5-24, Table
5-25,Table 5-26). Upstream, downstream, and total program impact are presented.
Table 5-24 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
-111,867,639
302.0
134.9
-9,324
-125.2
-12.6
2030
-219,811,320
631.8
284.1
-18,189
-245.7
-24.7
2040
-289,887,109
853.1
383.9
-21,642
-320.7
-32.6
2050
-368,990,880
1,087.4
489.6
-23,899
-403.0
-41.5
Table 5-25 Upstream GHG Change vs. Reference Case
A CO2 (Metric Tons)
A CH4 (Metric tons)
A N2O (Metric tons)
A GHG (MMT CO2 EQ)
2020
-27,200,175.2
-154,246.0
-437.2
-31.2
2030
-53,446,255.6
-303,081.5
-859.1
-61.3
2040
-70,484,907.4
-399,703.8
-1,133.0
-80.8
2050
-89,718,677.3
-508,774.1
-1,442.2
-102.9
5-30
-------
Emissions Impacts
Table 5-26 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
-139,067,814.2
-153,944.0
-302.3
-9,324.1
-156.3
-12.6
2030
-273,257,576.1
-302,449.7
-575.0
-18,189.3
-307.0
-24.7
2040
-360,372,016.8
-398,850.7
-749.1
-21,641.7
-401.5
-32.6
2050
-458,709,557.6
-507,686.7
-952.6
-23,899.2
-505.9
-41.5
5.5 Non-Greenhouse Gas Emission Inventory
The reference case emission inventories used for this 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.
The onroad mobile source numbers have been updated from the NPRM with the
MOVES data produced for this final rule analysis. For onroad mobile sources, the MOVES
2010 model was used as described in Section 5.3.3.5. This model estimates emissions from
light-duty and heavy-duty gasoline and diesel vehicles. These inventories have previously
been shown in Section 5.3.3.5.1. In some cases, particularly VOC, CO and NOx, the change
from the MOBILE model to a MOVES based inventory has led to large changes in reference
inventories from the proposal. These changes are due to model updates rather than program
changes.
Most nonroad equipment was modeled with NONROAD2005d, 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).46'4? This version of NONROAD does not include the county specific detail that
is provided when NONROAD is run using NMIM. Some precision is lost using this method.
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.46'48
Locomotives and C1/C2 commercial marine vessel inventories come from the spring 2008
5-31
-------
Regulatory Impact Analysis
final rule, and the C3 commercial marine emission inventory is from the base case inventories
in the June 2009 proposed rule.
Table 5-27 and Table 5-28 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 5-27 2020 Reference Case Emissions by Sector (annual short tons)
Onroad
Gasoline
Onroad Diesel
Nonroad SIa
Other
Nonroadb
Stationary/ Area
Total
voc
1,290,008
203,977
1,289,918
234,870
8,740,057
11,758,830
CO
20,764,531
612,037
14,286,250
1,424,643
11,049,239
29,756,282
NOX
1,493,306
1,502,844
242,828
3,389,761
5,773,927
30,783,084
S02
22,130
4,565
49,019
210,509
3,047,714
3,333,937
PM2.5
40,685
41,483
15,413
943,226
7,864,681
8,905,488
a Nonroad gasoline, LPG, and CNG engines plus portable fuel containers
' Nonroad diesel engines and all locomotive, aircraft, and commercial marine
TABLE 5-27
CONTINUED
Onroad
Gasoline
Onroad Diesel
Nonroad SIa
Other
Nonroadb
Stationary/ Area
Total
BENZENE
30,633
2,620
36,862
3,760
111,337
185,212
1,3-
BUTADIENE
4,777
1,397
5,895
929
1,847
25,038
ACETAL-
DEHYDE
10,805
5,562
4,768
9,542
13,118
33,602
FORMAL-
DEHYDE
11,268
15,279
10,240
22,324
23,846
82,957
ACROLEIN
569
741
584
1,013
3,412
6,319
' Nonroad gasoline, LPG, and CNG engines plus portable fuel containers
' Nonroad diesel engines and all locomotive, aircraft, and commercial marine
5-32
-------
Emissions Impacts
Table 5-28 2030 Reference Case Emissions by Sector (annual short tons)
Onroad
Gasoline
Onroad Diesel
Nonroad SIa
Other
Nonroadb
Stationary/ Area
Total
voc
1,006,387
209,384
1,198,679
238,652
8,740,057
11,393,159
CO
20,615,741
567,933
15,815,805
1,411,393
11,049,239
30,528,338
NOX
1,059,567
1,192,334
243,515
3,427,832
5,773,927
30,628,948
SO2
25,700
5,538
50,816
229,183
3,047,714
3,358,951
PM2.5
42,855
15,680
17,270
1,426,994
7,864,681
9,367,480
Table 5-28
continued
Onroad Gasoline
Onroad Diesel
Nonroad Sf
Other Nonroadb
Stationary/Area
Total
Benzene
22,048
2,689
39,871
3,764
111,337
179,709
1,3-
Butadiene
3,448
1,434
6,279
979
1,847
20,523
Acetal-
dehyde
7,909
5,709
5,118
9,579
13,118
30,683
Formal-
dehyde
8,196
15,684
11,229
22,487
23,846
81,442
Acrolein
419
761
629
1,055
3,412
6,276
1 Nonroad gasoline, LPG, and CNG engines plus portable fuel containers
' Nonroad diesel engines and all locomotive, aircraft, and commercial marine
5.5.1 Downstream Impacts of Program on Non-GHG Emissions
The non-GHG emission results shown here (Table 5-29) are a summary of the
previous analysis, and are combination of output from MOVES2010 and the OMEGA post-
processor.
5-33
-------
Regulatory Impact Analysis
Table 5-29 Downstream Emission Changes of Program
POLLUTANT
A 1,3 -Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Particulate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR YEAR
2020
Short Tons
-93.6
766.9
1.7
-750.0
10,156.3
2.1
13,410.3
231.0
-2,027.7
4,318.7
Percent
Change in
US Total
-0.37%
2.28%
0.03%
-0.40%
0.03%
0.00%
0.04%
0.00%
-0.06%
0.04%
CALENDAR YEAR
2030
Short Tons
-18.1
681.5
6.5
-248.3
182,787.8
116.3
16,142.6
602.3
-4,248.5
11,207.4
Percent
Change in
US Total
-0.09%
2.22%
0.10%
-0.14%
0.60%
0.14%
0.05%
0.01%
-0.13%
0.10%
5.5.2 Upstream Impacts of Program on Non-GHG Emissions
Non-GHG fuel production and distribution emission impacts of the 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) 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
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
5-34
-------
Emissions Impacts
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 joint TSD for today's rale.
Results of these emission inventory impact calculations relative to the reference case
for 2020 and 2030 are shown in Table 5-30 for the criteria pollutants and individual air toxic
pollutants.
The 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 5-30 Upstream Emission Changes of Program
POLLUTANT
A 1,3-Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Particulate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR YEAR
2020
Short Tons
-1.5
-6.8
-0.9
-139.6
-6,164.6
-51.4
-19,291.0
-2,629.1
-11,804.1
-64,505.9
Percent
Change in
US Total
-0.01%
-0.02%
-0.01%
-0.08%
-0.02%
-0.06%
-0.06%
-0.03%
-0.35%
-0.55%
CALENDAR YEAR
2030
Short
Tons
-3.0
-13.4
-1.8
-274.3
-12,113.0
-101.0
-37,905.4
-5,165.9
-23,194.1
-126,749.1
Percent
Change in
US Total
-0.01%
-0.04%
-0.03%
-0.15%
-0.04%
-0.12%
-0.12%
-0.06%
-0.69%
-1.11%
5.5.3 Total non-GHG Program Impact
Table 5-31 shows the combined impacts of downstream and upstream aspects of the
program. The net impacts of the program on VOC, NOx, and PM2.5, 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
emissions are driven by the rebound effect on VMT, which are only partially offset by
upstream reductions.
Net emissions depend on the relative impacts of the reductions from upstream
emissions versus increases due to the rebound effect and ethanol volume assumptions (that are
not due to the GHG vehicle standards) on the downstream emissions. All changes in non-
5-35
-------
Regulatory Impact Analysis
GHG emissions are less than 2.5% of the national inventory, with the net impact on most non-
GHG emissions at less than a single percent.
Table 5-31 Total Non-GHG Emission Changes of Program
POLLUTANT
A 1,3-Butadiene
A Acetaldehyde
A Acrolein
A Benzene
A Carbon Monoxide
A Formaldehyde
A Oxides of Nitrogen
A Particulate Matter
(below 2.5 micrometers)
A Oxides of Sulfur
A Volatile Organic Compounds
CALENDAR YEAR
2020
Short Tons
-95.1
760.0
0.8
-889.6
3,991.6
-49.3
-5,880.7
-2,398.1
-13,831.8
-60,187.1
Percent
Change in
US Total
-0.38%
2.26%
0.01%
-0.48%
0.01%
-0.06%
-0.02%
-0.03%
-0.42%
-0.51%
CALENDAR YEAR
2030
Short
Tons
-21.1
668.1
4.7
-522.5
170,674.8
15.3
-21,762.8
-4,563.6
-27,442.5
-115,541.7
Percent
Change in
US Total
-0.10%
2.18%
0.07%
-0.29%
0.56%
0.02%
-0.07%
-0.05%
-0.82%
-1.01%
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 joint TSD chapter 4. 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 5-32).
The ethanol volumes used in this analysis are from AEO 2007. As there are
proportionally few vehicles subject to the new GHG standards in the first few years of these
vehicle's lifetimes, which is when the majority of driving occurs, little change is anticipated
in the fuel supply which these vehicles use. Therefore, no fuel effects are calculated in the
MY lifetime analysis.
5-36
-------
Emissions Impacts
All other inputs, including sales and achieved emission levels are the same between
the two analyses.
Table 5-32 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
17,270
16,943
16,599
16,163
15,761
15,337
14,881
14,429
13,940
13,495
12,964
12,510
11,990
11,470
10,997
10,543
10,125
9,714
9,307
8,891
8,546
8,285
8,136
7,896
7,699
7,530
7,432
7,297
7,198
7,138
7,136
7,133
7,128
7,117
7,103
7,086
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
19,219
18,782
18,419
17,946
17,502
16,952
16,439
15,829
15,218
14,648
13,992
13,450
12,832
12,212
11,600
11,069
10,617
10,125
9,650
9,238
8,882
8,667
8,400
8,395
8,197
8,188
8,218
8,216
8,213
8,211
8,210
8,208
8,203
8,196
8,182
8,167
5-37
-------
Regulatory Impact Analysis
5.6.2 Results
The GHG emission reductions are shown for each model year, as are the co-pollutant
impacts (Table 5-33, Table 5-34).
Table 5-33 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 CO2
A Upstream CH4
AUpstream N2O
Total A Upstream
Total Program A GHG
Emissions
A Fuel Consumption
(Billion Barrels)
MY 2012
-59.1
-5.5
-6.6
0.0
0.0
-71.2
-15.7
-1.9
-0.1
-17.7
-88.8
-0.17
MY 2013
-84.0
-9.4
-11.2
0.0
0.0
-104.6
-22.7
-2.7
-0.1
-25.5
-130.2
-0.25
MY 2014
-108.9
-14.4
-17.2
0.0
0.1
-140.5
-30.0
-3.6
-0.1
-33.7
-174.2
-0.33
MY 2015
-153.9
-19.6
-23.4
0.0
0.1
-196.8
-42.2
-5.0
-0.2
-47.4
-244.2
-0.46
MY 2016
-214.5
-20.9
-25.0
0.0
0.1
-260.3
-57.2
-6.8
-0.3
-64.3
-324.6
-0.63
Program Total
-620.4
-69.9
-83.4
0.1
0.3
-773.4
-167.8
-20.0
-0.8
-188.7
-962.0
-1.85
5-38
-------
Emissions Impacts
Table 5-34 Lifetime non-GHG Emissions vs. Reference Case (Short Tons)
Downstream
AVOC
ANOX
A PM 2.5
AGO
ASO2
A Benzene
A 1,3 Butdiene
A Formaldehyde
A Acetaldehyde
A Acrolein
Upstream
AVOC
ANOX
A PM 2.5
AGO
ASO2
A Benzene
A 1,3 Butdiene
A Formaldehyde
A Acetaldehyde
A Acrolein
Total
AVOC
ANOX
A PM 2.5
AGO
ASO2
A Benzene
A 1,3 Butdiene
A Formaldehyde
A Acetaldehyde
A Acrolein
MY 2012
2,246.9
4,334.4
202.8
86,963.0
-1,362.3
76.1
12.8
30.7
29.2
1.4
-41,066.1
-12,281.2
-1,673.7
-3,924.6
-7,514.8
-88.9
-1.0
-32.7
-4.4
-0.6
-38,819.2
-7,946.8
-1,470.9
83,038.4
-8,877.1
-12.7
11.8
-2.1
24.9
0.8
MY 2013
3,281.5
6,414.4
296.7
127,079.2
-1,970.4
111.2
18.7
44.8
42.6
2.1
-59,394.3
-17,762.4
-2,420.7
-5,676.1
-10,868.7
-128.5
-1.4
-47.3
-6.3
-0.9
-56,112.8
-11,348.0
-2,124.1
121,403.1
-12,839.0
-17.4
17.3
-2.6
36.3
1.2
MY 2014
4,336.4
8,465.9
392.0
167,922.1
-2,601.0
146.9
24.7
59.2
56.3
2.7
-78,404.6
-23,447.5
-3,195.5
-7,492.9
-14,341.4
-169.7
-1.8
-62.5
-8.3
-1.1
-74,068.1
-14,981.6
-2,803.5
160,429.2
-16,948.4
-22.8
22.9
-3.3
48.0
1.6
MY 2015
6,055.7
11,745.4
547.0
234,434.9
-3,657.6
205.1
34.5
82.7
78.7
3.8
-110,253.7
-32,972.3
-4,493.6
-10,536.6
-20,175.5
-238.6
-2.6
-87.9
-11.7
-1.6
-104,197.9
-21,226.9
-3,946.6
223,898.3
-23,833.1
-33.4
31.9
-5.2
67.0
2.2
MY 2016
8,181.5
15,872.5
739.1
316,325.0
-4,964.0
276.7
46.5
111.5
106.2
5.2
-149,633.5
-44,749.1
-6,098.6
-14,300.0
-27,381.7
-323.8
-3.5
-119.3
-15.9
-2.2
-141,452.0
-28,876.7
-5,359.6
302,025.0
-32,345.7
-47.1
43.0
-7.8
90.3
3.0
Total
24,102.0
46,832.5
2,177.6
932,724.2
-14,555.2
816.0
137.2
328.9
313.1
15.3
-438,752.1
-131,212.5
-17,882.2
-41,930.1
-80,288.1
-949.4
-10.3
-349.8
-46.5
-6.4
-414,650.1
-84,379.9
-15,704.6
890,794.0
-94,843.3
-133.4
126.9
-20.9
266.6
8.9
5-39
-------
Regulatory Impact Analysis
5.7 Alternative 4% and 6% Scenarios
For this final rale, two alternative control scenarios were evaluated characterized by
4% and 6% annual growth in the GHG standards from the MY 2011 standard. Like the
previous analyses, this analysis has been updated from the proposal using the new economic
inputs. Other than the standards, these scenarios share all inputs with the 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 5-35 and the achieved levels are shown in Table
5-36.
Table 5-35 4% Scenario Standards
MODEL
YEAR
2012
2013
2014
2015
2016
PA
EMISSION
LEVEL
277
267
257
248
239
NPA EMISSION
LEVEL
365
352
329
327
315
ANTICIPATED
MY EMISSION
LEVEL
311
299
287
276
265
Table 5-36 4% Scenario Achieved Levels
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
277
269
261
251
239
ANTICIPATED
NPA EMISSION
LEVEL
380
367
349
335
315
ANTICIPATED
MY EMISSION
LEVEL
317
306
293
281
265
5.7.1.2 Results
Results are shown relative to the same reference scenario as the EPA program. Both
calendar year and model year lifetime results are shown.
5-40
-------
Emissions Impacts
Table 5-37 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)
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)
CY
2020
-89.8
-10.6
-13.3
0.0
0.0
-113.9
-24.5
-3.5
-0.1
-28.0
-141.9
-11.3
CY
2030
-183.4
-20.1
-26.0
0.0
0.0
-229.5
-49.5
-7.0
-0.2
-56.7
-286.2
-22.9
CY
2040
-243.0
-26.5
-30.9
0.0
0.1
-300.3
-65.5
-9.3
-0.3
-75.1
-375.4
-30.3
CY
2050
-309.5
-33.7
-34.2
0.0
0.1
-377.2
-83.5
-11.8
-0.3
-95.7
-472.9
-38.6
5-41
-------
Regulatory Impact Analysis
Table 5-38 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 (Billion gallons)
MY
2012
-20.7
-5.5
-6.6
0.0
0.0
-32.8
-6.4
-0.8
0.0
-7.2
-39.9
-2.9
MY
2013
-53.3
-9.4
-11.2
0.0
0.0
-73.9
-15.2
-1.8
-0.1
-17.1
-100.0
-7.1
MY
2014
-94.2
-14.4
-17.2
0.0
0.0
-125.7
-26.4
-3.1
-0.1
-29.7
-155.4
-12.2
MY
2015
-140.0
-19.6
-23.4
0.0
0.1
-182.9
-38.8
-4.6
-0.2
-43.6
-226.5
-18.0
MY
2016
-198.0
-20.9
-25.0
0.0
0.1
-243.8
-53.2
-6.3
-0.3
-59.8
-303.6
-24.6
Program
Total
-506.2
-69.8
-83.3
0.0
0.2
-659.0
-140.1
-16.7
-0.7
-157.4
-816.4
-64.8
5.7.2 6% Scenario
5.7.2.1 Standards and Achieved Levels
The program standards are shown in Table 5-35 and the achieved levels are shown in
Table 5-36.
Table 5-39 6% Scenario Standards
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
272
257
243
230
217
ANTICIPATED
NPA EMISSION
LEVEL
358
339
320
303
286
ANTICIPATED
MY EMISSION
LEVEL
305
288
272
256
241
5-42
-------
Emissions Impacts
Table 5-40 6% Scenario Achieved Levels
MODEL
YEAR
2012
2013
2014
2015
2016
ANTICIPATED
PA EMISSION
LEVEL
274
261
247
232
217
ANTICIPATED
NPA EMISSION
LEVEL
374
350
329
310
286
ANTICIPATED
MY EMISSION
LEVEL
313
295
277
260
241
5.7.2.2 Results
Results are shown relative to the same reference scenario as the EPA program. Both
calendar year and model year lifetime results are shown.
Table 5-41 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)
A N2O (MMT CO2 EQ)
A Total GHG
Total
A Total GHG
A Fuel Consumption (Annual, Billion gallons)
CY
2020
-137.4
-10.7
-13.3
0.0
0.0
-161.4
-36.0
-5.1
-0.2
-41.3
-202.7
-16.7
CY
2030
-274.9
-20.3
-26.0
0.0
0.1
-321.1
-71.8
-10.2
-0.3
-82.3
-403.4
-33.2
CY
2040
-363.2
-26.7
-30.9
0.0
0.1
-420.7
-94.8
-13.4
-0.5
-108.7
-529.4
-43.9
CY
2050
-462.4
-34.0
-34.2
0.0
0.2
-530.4
-120.7
-17.1
-0.6
-138.4
-668.8
-55.9
5-43
-------
Regulatory Impact Analysis
Table 5-42 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 (Billion gallons)
MY
2012
-37.7
-5.5
-6.6
0.0
0.0
-49.8
-10.5
-1.3
-0.1
-11.8
-61.7
-4.9
MY
2013
-96.9
-9.4
-11.2
0.0
0.0
-117.5
-25.8
-3.1
-0.1
-29.0
-146.5
-12.0
MY
2014
-158.1
-14.5
-17.3
0.0
0.1
-189.8
-42.0
-5.0
-0.2
-47.2
-237.0
-19.4
MY
2015
-222.8
-19.7
-23.5
0.0
0.1
-265.9
-59.0
-7.0
-0.3
-66.3
-332.2
-27.3
MY
2016
-295.1
-21.1
-25.1
0.0
0.1
-341.2
-76.9
-9.2
-0.4
-86.4
-427.6
-35.6
Program
Total
-810.6
-70.2
-83.8
0.1
0.4
-964.2
-214.2
-25.5
-1.1
-240.7
-1,204.9
-99.1
5.8 Inventories Used for Air Quality Analyses
This section describes the processes used in calculating the inventories for the air quality
(AQ) modeling analysis. Air quality modeling requires significant lead time, and
consequently the air quality inventories were completed significantly before the inventories
presented in this final rule.
5.8.1 Upstream Emissions
5.8.1.1 Petroleum Production and Refining Emissions
Petroleum production includes crude oil extraction and transport to refineries. As in
the nationwide analysis presented in the proposed rule as well as this final rule, we assumed
that (a) 50% of the change in gasoline supply was projected to come from domestic refineries,
and (b) 10% of the change in crude being used by domestic refineries would be domestic
crude. Thus, using our assumption that 1.0 gallon less of gasoline equates to approximately
1.0 gallon less crude throughput, the reduction in crude extraction and transport would equal
about 5% of the change in gasoline volume. To generate the emission inventory adjustment
factors for air quality modeling these reductions were applied to the projected crude supply to
US refineries, per AEO 2009 (stimulus version).49 The resulting estimates are shown in Table
5-43. Only the 2030 values were used in the air quality modeling. The percent reductions
were applied to the NEI projected inventories for 2030. The 0.61% reduction was applied to
5-44
-------
Emissions Impacts
all SCCs associated with petroleum extraction, and the 6.09% reduction was applied to all
SCCs associated with gasoline refining.50
Table 5-43 Crude Oil and Gasoline Volume Reductions Associated with LD GHG Rule
PARAMETER
Crude Supply to US Refineries (bgal/yrf
Reduction in Gasoline Consumption (bgal/yr)
Reduction in Domestic-Refined Gasoline (bgal/yr)
Reduction in Domestic Refining of Crude (US & Imported
Crude) (bgal/yr)
Reduction in Domestic-Refined Gasoline from Domestic Crude
(bgal/yr)
Reduction in Domestic Crude Production & Transport to
refineries (bgal/yr)
Percent Reduction in Domestic Refining
Percent Reduction in Domestic Crude Production & Transport
2020
211.96
13.35
6.68
6.68
0.67
0.67
3.17%
0.32%
2030
214.94
26.18
13.09
13.09
1.31
1.31
6.09%
0.61%
Note that this method used for AQ county allocation is not directly comparable with
the method used for nationwide impacts, for which we used the GREET-based upstream
impacts spreadsheet model to calculate the absolute change in tons for each stage of the
upstream inventory.
5.8.1.2 Gasoline Transport, Storage and Distribution emissions (vapor)
With the reduced gasoline consumption associated with this rule there would be
changes in the quantity of vapor losses during the transport and distribution of gasoline. The
analysis of these impacts was separated into two segments: refinery to bulk terminal (RBT)
and bulk terminal to pump (BTP). The reference case analyzed would include some amount
of EO (zero percent ethanol) in the BTP segment, but the reduced gasoline production
projected with this rule, combined with unchanged ethanol volumes, means that essentially all
gasoline would be blended with at least 10 percent ethanol. Thus the transport of EO gasoline
would only occur between refineries and blending terminals in the control case, i.e., the RBT
segment. The BTP segment would include both E10 and E85. No changes in volumes of E10
or E85 were assumed for this analysis.
EO - Refinery to Bulk Terminal (RBT)
EO - Bulk Terminal to Pump (BTP, used for reference cases only)
E10 - Bulk Terminal to Pump (BTP)
E85 - Bulk Terminal to Pump (BTP)
For each of the above fuel type and transport stage combinations, nationwide VOC
impacts (ton deltas) (and benzene and ethanol vapor) were calculated using EPA's upstream
impacts spreadsheet model for the control scenario versus the reference case. For air quality
5-45
-------
Regulatory Impact Analysis
modeling the three BTP values were combined into a total BTP impact for each scenario.
These impact values were renormalized to be ton deltas relative to the reference case. Then
all the SCCs in the NEI related to gasoline transport, storage, and distribution (TS&D) were
categorized as either RBT or BTP, and the NEI VOC emissions were summed for each
category. The nationwide VOC percent change for the control case relative to the reference
case for RBT was calculated as the control case RBT delta tons (versus AEO) divided by the
NEI RBT tons. Similarly, the nationwide VOC percent change for the control case for BTP
was calculated as the control case BTP delta tons (versus AEO) divided by the NEI BTP tons.
The projected reference case VOC and calculated adjustments are shown in Table 5-44.
Table 5-44 Gasoline Distribution Emission Reductions Associated with LD GHG Rule
GASOLINE
DISTRIBUTION
SEGMENT
Refinery to Bulk
Terminal
Bulk Terminal to
Pump
FUEL
EO
EO
E10
E85
Total
NEI VOC
(TON/YR)
273,513
490,236
VOC
CHANGE
(TON/YR)
-50,042
-92,034
0
0
-92,034
PERCENT
CHANGE VS
NEI
-18.30%
-18.77%
The county level AQ inventories for the control case were then calculated by applying
these percent changes in VOC to the corresponding sets of SCCs (point and non-point
sources) for every county. The same adjustment factors were applied to benzene evaporative
emissions.51
5.8.1.3 Downstream Emissions from Onroad vehicles and Nonroad Equipment
except Aircraft, Locomotives, and Commercial Marine Vessels
5.8.1.4
Introduction
Downstream emissions are those resulting from mobile-source operation, including
onroad vehicles and nonroad vehicles and equipment. This section describes the development
of emissions from all onroad vehicles and from nonroad equipment modeled by the
NONROAD Model. Emissions from aircraft, locomotives, and commercial marine vessels
are discussed in Section 0.
The emissions discussed in this section were developed using three EPA models:
MOVES, MOBILE6, and NONROAD. MOBILE6 and NONROAD were run using the
National Mobile Inventory Model (NMIM), which is software that runs MOBILE6 and
NONROAD at the county-month level by accessing a county database, preparing input files,
and aggregating the output.
Similar to the emission inventories for the final rule, both VMT rebound and the
effects of increasing ethanol proportions in the fuel supply (which are not due to the GHG
vehicle standards) were accounted for in the air quality modeling inventories. Additional
5-46
-------
Emissions Impacts
details on the downstream inventories used for the air quality modeling are available in a
docket memo.52
5.8.1.5 Onroad
For onroad mobile sources except motorcycles, EPA executed an internal draft version
of MOVES dated 8/25/2009 (MOVES20090825), which is similar to Draft MOVES200953
with a number of improvements to the fuel effects data and code. This version of MOVES
used default database MOVESDB20090902. A slightly later version, MOVES20091113,
was run for evaporative runs to correct a bug in the code that processed evaporative
emissions. This version of MOVES used default database MOVESDB20091109. User-
supplied fuel and temperature tables were used for these runs. Historical temperature and
humidity data for 2005 was used for all years. For motorcycles, we relied on the MOBILE6.2
model as run using the NMIM platform with county-specific fuel properties and temperatures.
MOVES supplied all pollutants except SO2 and NH3, which came from NMIM runs. Onroad
inventories were generated by multiplying MOVES emission factors by VMT developed for
the Office of Air Quality Planning and Standards's 2002 Version 3 Modeling Platform
(PF02v3)54 and used in the recently published Locomotive-Marine Rule.55 This VMT, which
was based on AEO2006,56 was adjusted to match the annual VMT from AEO2009,57 but with
county allocations preserved. AEO2009 growth factors from 2005 to 2030 were applied to
2005 NEI VMT using the three AEO categories: light duty, commercial light trucks and
heavy duty. Assignments to MOBILE6 categories are straightforward except for Commercial
Light Trucks, which are only gasoline in the AEO classification. The MOBILE6 Model was
M6203ChcOxFixNMIM, a special version that includes cold-start VOC and the cold-start
controls of the Mobile Source Air Toxics Rule that go into effect in 2011 and used in the
recently published Locomotive-Marine rule. Onroad emissions generated at the state-month
level from MOVES were distributed to the county-month level using the results from
MOBILE6 as run by NMIM.
5.8.1.6 Nonroad
Nonroad equipment except for aircraft, locomotives, and commercial marine vessels
was modeled with the latest publically released NONROAD version NONROAD2008a.58
This version of the NONROAD includes the benefits of the two nonroad regulations
published in 2008 (the Locomotive and Marine Rule and the Small Spark-Ignition And
Recreational Marine Engine Rule59) plus all previous nonroad regulations.
5.8.1.7 Summaries
The two tables below are national-annual emissions in U.S. tons by mobile-source
sector for the reference and control cases in 2030. There are differences between these
inventories and those produced for the Final Rule because the latter used the final version of
MOVES2010 and an earlier version was used for the air quality modeling inventories. This
difference was unavoidable due to the long lead time required for air quality modeling.
However, the proportional difference in downstream inventories between the Control and
Reference cases is nearly the same in the two sets of inventories, so the air quality modeling
adequately reflects the effects of the rule.
5-47
-------
Regulatory Impact Analysis
Table 5-45 2030 Reference Case. National annual emissions in U.S. tons
pollutant
VOC
NOX
CO
SO2
PM2.5
Benz
Acet
Bute
Formal
Aero
Sector
Light duty
Gasoline
635,169
1,245,018
17,901,362
33,384
24,224
15,723
5,823
2,520
5,996
275
Light duty
Diesel
1,122
7,776
8,441
884
112
23
14
10
44
4
Heavy Duty
Gasoline
22,921
63,982
783,117
2,185
829
606
231
98
235
22
Heavy Duty
Diesel
167,574
979,298
477,617
3,633
16,258
1,843
5,054
1,070
13,723
614
Non-road
Diesel
67,377
481,698
165,956
1,314
18,817
1,371
3,576
125
7,961
204
Non-Road
Gasoline
1,170,391
207,374
15,148,540
1,104
48,012
23,952
3,284
3,924
6,188
358
Table 5-46 2030 Control Case. National annual emissions in U.S. tons.
pollutant
VOC
NOX
CO
SO2
PM2.5
Benz
Acet
Buta
Formal
Aero
Sector
Light duty
Gasoline
651,538
1,268,205
18,126,698
27,975
24,575
15,652
6,341
2,519
6,108
280
Light duty
Diesel
1,139
7,893
8,578
740
113
23
14
10
44
4
Heavy Duty
Gasoline
23,160
64,179
780,432
2,185
829
593
248
97
235
22
Heavy Duty
Diesel
167,574
979,298
477,617
3,633
16,258
1,843
5,054
1,070
13,723
614
Non-road
Diesel
67,377
481,698
165,956
1,314
18,817
1,371
3,576
125
7,961
204
Non-Road
Gasoline
1,170,521
216,640
14,706,415
1,106
48,012
23,009
3,470
3,974
6,258
352
5-48
-------
Emissions Impacts
5. A Appendix to Chapter 5: Details of the TLA AS Impacts Analysis
5.A.I 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. As a change from the proposed rule, the final rule allows that
manufacturers with total domestic sales of less than 50,000 in MY 2009 to place up to
250,000 vehicles from model years 2012-2016 into a separate fleetJ'K. 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.B.5 of the preamble). One commenter, the
American Council for an Energy Efficient Economy, voiced concerns that EPA (A)
underestimated the impact of the TLAAS program, and (B) did not provide documentation of
the relevant calculations. As in the proposal, EPA has provided documentation of the
calculation in this appendix, with the relevant spreadsheets available in the docket.60 EPA has
also revised its estimates of the TLAAS provision for this final rule. Please see the Response
to Comments document for additional details. Several manufacturer decisions and
marketplace events will ultimately 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 37 MMT of CO2
emissions, in all cases the TLAAS program has a proportionally small impact (< 4%) on the
total program benefits over the model years 2012-2016. The maximum impact presented here
has increased approximately proportional to the increase in program size from the proposal;
i.e., the maximum potential impact described in the NPRM was 25 MMT for 1.1 million
vehicles, while the maximum impact described here is 37 MMT for 1.55 million vehicles.
Under the estimation procedure used in the emission inventory analysis (as opposed to
the bracketing analysis mentioned immediately above), the TLAAS program is projected to
result in an approximately 14 MMT decrease in greenhouse gas benefits from this rule over
the lifetime of vehicles manufactured in model years 2012-2016 (assuming that it is
technically feasible for all TLAAS-eligible producers to meet the otherwise-applicable GHG
standards for those years, a dubious assumption given the very short lead times available).
While 14 MMT is a small fraction of the overall program benefits (approximately one
percent of the estimated GHG reductions), 14 MMT is an increase over the 3.4 MMT impact
JThese manufacturers could place up to 200,000 vehicles into the TLAAS fleet between MYs 2012-2015, and up
to an additional 50,000 vehicles into that fleet in MY 2016.
KThe final rule also deters regulation of manufacturers of vehicles with 2008 or 2009 domestic sales of less than
5,000 vehicles whose three-year rolling average of domestic sales remain less than 5,000 vehicles, and which
demonstrate inability to purchase credits. The deferral is respect to the CO2 standards only. These vehicles are
not considered in the analysis above.
5-49
-------
Regulatory Impact Analysis
estimated in the proposal. This is largely due to several changes in the TLAAS program made
in response to public comment, as discussed below. There was also a calculation error in the
analysis for the proposal, which would have increased the estimate of the TLAAS, as
proposed to 4.9 MMT.
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, Jaguar, Mazda, Mitsubishi, Suzuki, Daimler, Subaru, BMW,
Volkswagen, Hyundai, and Kia. Three of the above manufacturers are expected to be eligible
for the expanded TLAAS program. These manufacturers are Suzuki, Porsche and Jaguar.
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.
The impacts of the program therefore partially depend on manufacturer eligibility.
Manufacturer Usage
As explained in section III.B of the preamble to the final rule, the TLAAS program is
predicated on the need for additional lead time for certain manufacturers, and is a reasonable
exercise of EPA's section 202 (a) to consider lead time in crafting standards. The TLAAS
provides needed flexibility to manufacturers in order to comply with the CO2 standards for the
earlier model years,, thereby providing needed lead time for these manufacturers to bring
their entire fleet into compliance with the stringent 2016 MY standards or, for manufacturers
eligible for TLAAS in MY 2016, to be fully compliant with the standards by the 2017 model
year. 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.
Further, when the TLAAS program ends, manufacturers' entire fleets must 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.
5-50
-------
Emissions Impacts
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 in
estimating its impacts. 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 which 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 CCh emission standard. 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, plus 250,000 vehicles over a five-year period for eligible smaller manufacturers
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 976 MMT GHG reduced over
lifetime of model years 2012-2016 (i.e. impact of the TLAAS is zero), while the upper bound
impact was 938 MMT GHG reduced over the same period. Thus, the maximum potential
impact of the program, even under this most extreme scenario is approximately 37 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 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 RIA chapter 5. In this analysis, all eight TLAAS manufacturers were assumed to
use the example vehicle allocation schedule from Section III of the proposal Preamble,
5-51
-------
Regulatory Impact Analysis
replicated in Table 5-47. 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 5-47 TLAAS Default Vehicle Production Volumes
MODEL
YEAR
Sales Volume
2012
40,000
2013
30,000
2014
20,000
2015
10,000
The allocation was split evenly between cars and trucks for each manufacturer. For
these eight manufacturers, the TLAAS fleet allotment was assumed to emit as much COi per
mile as expected from the largest complying footprint car or truck in each manufacturer's
fleet. For emission estimation purposes, upsizing the fleet effectively lowers the stringency of
the target. 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 emission level.
For the three manufacturers eligible for the expanded TLAAS, the phase-in schedule
shown in the table below was used. As this allocation encompasses almost all of the
manufacturers' sales in the early years, a different allocation was used for each manufacturer.
Almost all of Porsche's fleet, both cars and trucks, is covered. Tata is assumed to split the
allocation between car and trucks. Suzuki is assumed to use the entire allocation for its cars.
As the compliance gap for the smaller manufacturers is on average significantly larger
than the average compliance gap for the other 8 manufacturers, these manufacturers were
assumed to make a more intense use of the TLAAS program. For these manufacturers, their
TLAAS fleets are assumed to emit 1.25x more emissions than the manufacturer's sales
weighted target in 2012, and by 2016, they were assumed to emit 1.05x more emissions. This
schedule assumes a gradual decrease in CO2 emissions, which would project that the
manufacturers to reach complicance with the main program CO2 standards by 2017. The
volumes assumes slight growth from the 2009 base year and are based on the EPA fleet data
file. There is uncertainty in these estimates, as discussed above.
Table 5-48 Volumes and Usage Ratios in Expanded TLAAS Analysis
MODEL
YEAR
Sales
Volume
Usage Ratio
2012
60,000
1.25
2013
60,000
1.20
2014
50,000
1.15
2015
40,000
1.10
2016
40,000
1.05
5-52
-------
Emissions Impacts
The expanded TLAAS program grows the program by approximately 40% (1.55
million vehicles now /1.1 million vehicles in the proposal). The increased program size,
combined with the assumption that the users of the expanded TLAAS program will use it
more heavily accounts for the increased estimate of impacts. As in the proposal, small
volume manufacturers are not included in this final rule analysis.
In this analysis, the total TLAAS program results in an emission impact of
approximately 14 MMT CO2 over the lifetime of the 2012-2016 MY vehicles.
The gram per mile impacts are listed here for each of these scenarios.
Table 5-49 Gram per Mile per Year
Model Year
2012
2013
2014
2015
2016
TLAAS impact (Grams CO2
Emissions Per Mile)
Lower
Bound
Scenario
0.0
0.0
0.0
0.0
0.0
Upper
Bound
Scenario
2.6
2.1
1.6
1.0
0.5
Estimate
Used In
Emission
Analysis
1.2
0.9
0.6
0.3
0.1
5-53
-------
Regulatory Impact Analysis
5.B Appendix to Chapter 5: Impacts of Advanced Technology Vehicle
Incentives for Electric Vehicles
5.B.1 Introduction and Summary
As described in Preamble Section III, EPA is finalizing provisions that provide a
temporary regulatory incentive for the commercialization of certain advanced vehicle power
trains—electric vehicles (EVs), plug-in hybrid electric vehicles (PHEVs), and fuel cell
vehicles (FCVs)—for model year 2012-2016 light-duty and medium-duty passenger vehicles.
EPA is finalizing two changes to the proposed incentive program—deleting the vehicle
multiplier and adding an individual automaker cap on the cumulative vehicle production
eligible for the zero gm/mi compliance value —that will limit the loss in GHG savings due to
these incentives. These incentives apply for the model years 2012-2016 covered by this final
rule, and EPA will revisit this issue in rulemakings for future model years.
This section provides an analysis of the emission impacts of 500,000 electric vehicles
produced under the zero gram per mile incentive during the 2012-2016 timeframe. As stated
in Preamble Section III, it is impossible to predict the number of EVs that will be produced
between 2012 and 2016. EPA believes that sales of 500,000 "un-capped" EVs is a reasonable
scenario. Fewer EVs, or a combination of 500,000 EVs and PHEVs, would lessen the loss in
GHG benefits. Conversely, additional sales of "un-capped" EVs would increase the loss in
GHG benefits.
Based on the analysis presented here, sales of 500,000 uncapped electric vehicles
would produce a net reduction in GHG benefits of 25 MMT over the program without this
provision.
5.B.2 Assumptions behind the Analysis
This analysis is intended as a preliminary exercise in an emerging field. The net
impacts of the EV provision are dependent on several assumptions, and the assessment
published here is intended to be demonstrative. Several assumptions are conservative; most
significantly, the assumptions regarding manufacturer usage of the EV provision are meant to
be a boundary analysis.
We assume that manufacturers utilize the full benefit of the EV provision to
manufacture internal combustion engine vehicles that emit more than they would otherwise.
As an example, the fleet target of a typical manufacturer would be 250 grams CC>2 per mile in
2016. If the manufacturer sold EVs under this program, the manufacturer could place less
technology on their conventional vehicles, so long as their net achieved level met their target.
In essence, the downstream (tailpipe) benefit of the EV would be canceled out. Because the
same amount of gasoline would be consumed regardless of the EV provision, sales of EVs
would therefore not impact total fuel savings or total emissions from gasoline vehicles.
Upstream emissions for electric vehicles sold beyond the individual automaker cap would be
accounted for by this rule, and would therefore have no negative environmental impact.
5-54
-------
Emissions Impacts
We assume 500,000 electric vehicles are sold in the 2012-2016 timeframe, which is a
significant increase over current electric vehicle sales, which are near zero in EPA's 2008
baseline market data file. Similarly, EPA's OMEGA modeling does not predict the need for
significant electrification of the fleet over the time frame presented in this rulemaking.
We assume that EVs are only sold as cars, and that they are driven for the same
lifetime VMT as a conventional vehicle.
5.B.3 Inputs
As stated previously, we assume that there is no downstream benefit of electric
vehicles, and therefore the net impact of electric vehicles is equivalent to the net impact of the
electricity generated to fuel these vehicles. To calculate the net impact over the vehicle
lifetime of all the EVs sold, the formula is as follows:
EV emission impact = (Lifetime VMT) x (kWh/mile) x (Emissions/kWh) x (sales)
Equation 6- EV Impacts
The inputs used for this calculation are shown below (Table 5-51). Given that
Equation 6 is linear, a change to any of these variables would produce a proportional and
linear change in the results of the analysis.
Table 5-50 Inputs for EV Analysis
LIFETIME
VMT
kWh/mile
GHG Emissions
g CO2e/kWh
195,264
0.329
768
The lifetime VMT is described in TSD Chapter 4 and in Table 5-32, and is assumed to
be the same between conventional vehicles and electric vehicles.
The kWh per mile value is intended as a rough estimate of the potential electricity
usage of typical mid-size EV in the 2012-2016 timeframe. Based on preliminary EPA
analysis, an EV of approximately 3200 pounds is projected to consume 230 Wh/mile over the
combined FTP and HWFE tests. It is assumed that on-road energy consumption will be about
5-55
-------
Regulatory Impact Analysis
43% higher than the tested energy consumption1; Based on these assumptions, we project a
mid-size EV would have a real world electrical energy consumption rate of 329 Wh/mile.
Table 5-51 EV energy required
FTP/HWFE
Fuel Energy
required
FTP/HWFE
Real World
Energy Required
230
329
Accounting for the CC^e emissions from EVs requires accounting for emissions
during the feedstock gathering, power generation, power distribution, and vehicle charging
stages. In other words, accounting for electricity for EVs requires accounting for the
efficiencies of the various stages of the combustion of fossil fuels at the power plant, as well
as the inefficiencies in transmitting the electricity from plant to the vehicle. For this analysis,
the electricity is generated at the 2005 national average emission level (633 g CCV kWh).61M
This value must be adjusted for emissions due to charging inefficiencies ("wall to
vehicle" losses) of 10%, and transmission and distribution losses ("plant to wall") of 7%,
resulting in an actual upstream emission impact of 768 g CC^e for each kWh used at the
vehicle.
The emission factors used in this analysis could be higher or lower depending on
when users charge their vehicles, and whether this causes additional natural gas or coal power
plants to be shunted on-line. If the CC^e emissions from powerplants were 10% higher or
lower, the resulting impacts from the EV provision would be proportionally higher or lower.
5.B.4 Computation
The inputs from Table 5-51 were inserted in Equation 6, with the resulting impact
shown below.
L Based on preliminary data, we assume that on-road EV shortfall is greater for electric vehicles than typical ICE
vehicles. This accounts for performance in different climates, as well as other issues. We assume this shortfall
to be about 30%, if measured in terms of fuel economy. If converted to an increase in energy consumption, that
factor becomes 1/0.7 or a 43% increase.
M The value 633 g CO2/kWh was derived by beginning with the EPA eGrid 2007 v 1.1 emissions, combining
N2O, CH4 and CO2 using GWP values stated in Table 5-5, and adding 6% for feedstock gathering based on
Greet 1.8.
5-56
-------
Emissions Impacts
195,265 miles x 329 kWh/mile x 0.768 g CO2/kWh x 500,000 sales / (1012 conversion of grams to
MMT)
= 24.8MMT
References
1 Intergovernmental Panel on Climate Change Working Group I. 2007. Climate Change 2007
- The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change.
2 U.S. Environmental Protection Agency. 2009. Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1990-2007. EPA 430-R-09-004. Available at
http://epa.gov/climatechange/emissions/downloads09/GHG2007entire report-508.pdf
3 U.S. EPA. 2009 Technical Support Document for Endangerment and Cause or Contribute
Findings for Greenhouse Gases under Section 202(a) of the Clean Air Act. Washington, DC.
pp. 180-194. Available at
http://epa.gov/climatechange/endangerment/downloads/Endangerment%20TSD.pdf
4 U.S. Environmental Protection Agency. 2009. Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1990-2007. EPA 430-R-09-004. Available at
http://epa.gov/climatechange/emissions/downloads09/GHG2007entire report-508.pdf
5 RIA Chapter 2
6 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/. Docket ID: EPA-HQ-
OAR-2009-0472-0215
7 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
8 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 . Docket ID: EPA-HQ-OAR-2009-0472-01 17
9 EPA. MOVES 2010. http://www.epa.gov/otaq/models/moves/index.htm
10
U.S. EPA 2009. Updated OMEGA Post-Processor Spreadsheet. February 2010.
11
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
Docket ID: EPA-HQ-OAR-2009-0472-01 16
5-57
-------
Regulatory Impact Analysis
12 MOVES documentation and technical documents can be seen at
http://www.epa.gov/otaq/models/moves/index.htm.
13 EPA's OMEGA model, documentation, and technical documents can be found
http://www.epa.gov/oms/climate/models.htm. The model is also docketed EPA-HQ-OAR-
2009-0472-0192.
14 U.S. EPA 2010, Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis.
EPA-420-R-10-006. February 2010. Docket EPA-HQ-OAR-2009-0472-11332. Chapters 2
and 3.
15 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/. Docket ID: EPA-HQ-
OAR-2009-0472-0215
16 U.S. EPA. 2008. RFS2 Modified version of GREET1.7 Upstream Emissions Spreadsheet,
October 31,2008. Docket ID: EPA-HQ-OAR-2009-0472-0191
17 U.S. EPA, FRM Achieved CO2 standards worksheet. 2010.
18 NHTSA. 2009. Average Fuel Economy Standards, Passenger Cars and Light Trucks,
Model Year 2011. Docket ID: NHTSA-2009-0062-0001. Docket ID: EPA-HQ-OAR-
2009-0472-0060
http://www.nhtsa.dot.gov/portal/nhtsa_static_file_downloader.jsp?file=/staticfiles/DOT/NHT
SA/Rulemaking/Rules/Associated Files/CAFE_Updated_Final_Rule_MY2011 .pdf
19 When this rule's analysis was initiated, the RFS2 rule was not yet final. Therefore, it
assumes the ethanol volumes in Annual Energy Outlook 2007 (U.S. Energy Information
Administration, Annual Energy Outlook 2007, Transportation Demand Sector Supplemental
Table, http://www.eia.doe.gov/oiaf/archive/aeo07/supplement/index.html)
20 EPA 2010, Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis. EPA-
420-R-10-006. February 2010. Docket EPA-HQ-OAR-2009-0472-11332. see also 75 FR
14670, March 26, 2010
21 EPA. 2009. Regulation of Fuels and Fuel Additives: Changes to Renewable Fuel Standard
Program. EPA-HQ-OAR-2005-0161. http://www.epa.gov/otaq/renewablefuels/rfs2 1-
5.pdf. Docket ID: EPA-HQ-OAR-2009-0472-0119
22 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/NHT
SA/Rulemaking/Rules/AssociatedFiles/CAFE Updated Final Rule MY2011.pdf. Docket
ID : EPA-HQ-OAR-2009-0472-0060
23 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. August
2009. EPA-HQ-OAR-2009-0472-0085.
5-58
-------
Emissions Impacts
24 Energy Information Administration. Annual Energy Outlook 2010 Early Release.
Supplemental Transportation Tables. December 2009.
http://www.eia.doe.gov/oiaf/aeo/supplement/sup tran.xls.
25 Energy Information Administration. Annual Energy Outlook 2009. Supplemental
Transportation Tables. April 2009.
http://www.eia.doe.gov/oiaf/aeo/supplement/sup tran.xls. EPA-HQ-OAR-2009-0472-0121
26U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. February
2009..
27 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/NHT
SA/Rulemaking/Rules/AssociatedFiles/CAFE Updated Final Rule MY2011.pdf. Docket
ID : EPA-HQ-OAR-2009-0472-0060
28 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/NHT
SA/Rulemaking/Rules/AssociatedFiles/CAFE Updated Final Rule MY2011.pdf. Docket
ID : EPA-HQ-OAR-2009-0472-0060
29 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. February
2009..
30 EPA. Emission Facts: Average Carbon Dioxide Emissions Resulting from Gasoline and
Diesel Fuel. EPA420-F-05-001 February 2005. Docket ID: EPA-HQ-OAR-2009-0472-0122
31 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. February
2009.
32 Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy.
Transportation Energy Data Book: Edition 27. Chapter 4. 2008. Docket ID: EPA-HQ-
OAR-2009-0472-0280.
33 NHTSA. Vehicle Survivability and Travel Mileage Schedules. 2006. Docket ID: EPA-
HQ-OAR-2009-0472-0126
34 Joint TSD Chapter 4
35 NHTSA. Vehicle Survivability and Travel Mileage Schedules. 2006. Docket ID: EPA-
HQ-OAR-2009-0472-0126
36 Joint TSD Chapter 4
5-59
-------
Regulatory Impact Analysis
37 EPA. Final Technical Support Document.Fuel Economy Labeling of Motor Vehicle
Revisions to Improve Calculation of Fuel Economy Estimates. EPA-HQ-OAR-2009-0472-
0281
38 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.
Docket ID: EPA-HQ-OAR-2009-0472-0116
39 RIA chapter 2.
40 Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy.
Transportation Energy Data Book: Edition 27. 2008. Docket ID: EPA-HQ-OAR-2009-0472-
0280.
41 EPA. Final Technical Support Document. Fuel Economy Labeling of Motor Vehicle
Revisions to Improve Calculation of Fuel Economy Estimates. EPA-HQ-OAR-2009-0472-
0281
42 EPA. Emission Facts: Average Carbon Dioxide Emissions Resulting from Gasoline and
Diesel Fuel. EPA420-F-05-001 February 2005. Docket ID: EPA-HQ-OAR-2009-0472-0122
43 EPA. Technical Description of the Toxics Module for MOBILE6.2 and Guidance on Its
use for Emission Inventory Preparation. November 2002.
44 U.S. EPA 2010, Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis.
EPA-420-R-10-006. February 2010. Docket EPA-HQ-OAR-2009-0472-11332. Chapters 2
and 3.
45 Craig Harvey, EPA, "Calculation of Upstream Emissions for the GHG Vehicle Rule."
2009. Docket ID: EPA-HQ-OAR-2009-0472-0216
46 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). Docket ID: EPA-HQ-OAR-
2009-0472-0139
47 Control of Emissions From Nonroad Spark-Ignition Engines and Equipment, Final Rule
(Federal Register Vol 73, No. 196, page 59034, October 8, 2008). Docket ID: EPA-HQ-
OAR-2009-0472-0282
48 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/. Docket ID: EPA-HQ-OAR-2009-0472-0283
49 Energy Information Administration. Annual Energy Outlook 2009. Supplemental
Transportation Tables. April 2009.
http://www.eia.doe.gov/oiaf/aeo/supplement/sup tran.xls. EPA-HQ-OAR-2009-0472-0121
5-60
-------
Emissions Impacts
50 U.S. EPA. 2009. "For LD GHG AQ Modeling, 2030 Control Case: Adjustments to Oil
Refining, and Crude Production/Transport SCCs," spreadsheet file:
"Oil_Prod_Transp_Refine_2030_adjust.xls," 11/6/2009.
51 U.S. EPA. 2009. "Gasoline Distribution SCCs and Adjustments to apply for the LD GHG
AQ Modeling 2030 Control Case," spreadsheet file: "LDGHG_SCC_GasDistrib_Adjust2.xls"
11/9/2009.
52 Harvey Michaels, US EPA. "NMIM and MOVES Runs for LD GHG Air Quality
Modeling: Memorandum" 3/18/2010. Accompanying physical DVD which contains relevant
software and processing scripts. Title: "NMIM and MOVES Runs for LGR Air Quality
Modeling: DVD".
53 For information on Draft MOVES 2009, see
http://www.epa.gov/otaq/models/moves/movesback.htm.
54 See http://www.epa.gov/ttn/chief/emch/index.htmltt2002.
55 Final Rule: Control of Emissions of Air Pollution from Locomotives and Marine
Compression-Ignition Engines Less Than 30 Liters per Cylinder (published May 6, 2008 and
republished June 30, 2008). For details, see
http://www.epa.gov/otaq/locomotives.htmtt2008final.
56 Energy Information Administration. Annual Energy Outlook 2009. See
http://www.eia.doe.gov/oiaf/archive/aeo06/index.htmle
57 Energy Information Administration. Annual Energy Outlook 2009. See
http://www.eia.doe.gov/oiaf/archive/aeo09/index.html
58 NONROAD Model 2008a. http://www.epa.gov/otaq/nonrdmdl.htm
59 Final Rule: Control of Emissions from Nonroad Spark-Ignition Engines and Equipment
(published October 8, 2008). For details, see http://www.epa.gov/otaq/equip-ld.htm. Docket
ID: EPA-HQ-OAR-2009-0472-0282
60 U.S. EPA, FRM Achieved CO2 standards worksheet. 2010.
61 U.S. EPA. 2009. eGrid2007 dataset. http://www.epa.gov/cleanenergv/energy-
resources/egrid/index.html. Accessed February 3, 2010.
5-61
-------
Vehicle Program Costs Including Fuel Consumption Impacts
CHAPTER 6: Vehicle Program Costs Including Fuel
Consumption Impacts
This chapter presents the costs of the 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 rule. 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 rule.
Such impacts are presented in Chapter 8 of this RIA.
The costs presented here differ slightly from those presented in the proposal. The
different costs for the vehicle program are the result of revised costs for a limited set of
technologies expected to be used for compliance. Those revised costs stem from the
continuing teardown cost estimation work being done by FEV for EPA. See 74 FR at 48502.
At proposal, we used the tear down data to estimate the cost of stoichiometric gasoline direct
injection and turbocharging with engine downsizing. We have expanded our use of FEV tear
down costs for the final rule to estimate costs for the following technologies using new
information available to us shortly after the proposal: stoichiometric gasoline direct injection
and turbo charging with engine downsizing for a single overhead cam (SOHC) 3
valve/cylinder V8 engine downsized to a SOHC V6 engine; stoichiometric gasoline direct
injection and turbo chargin with engine downsizing for a dual overhead cam (DOHC) V6
engine to a DOHC 4 cylinder engine; a 6 speed automatic transmission replacing a 5 speed
automatic transmission; and a 6 speed wet dual clutch transmission replacing a 6 speed
automatic transmission.
This costing methodology has been published and gone through a peer review.1 In
addition, FEV and EPA extrapolated the engine downsizing costs for the following scenarios
that were outside of the noted study cases:2
1. Downsizing a SOHC 2 valve/cylinder V8 engine to a DOHC V6 engine.
2. Downsizing a DOHC V8 engine to a DOHC V6 engine.
3. Downsizing a SOHC V6 engine to a DOHC 4 cylinder engine.
4. Downsizing a DOHC 4 cylinder engine to a DOHC 3 cylinder engine.
For more detail on those revised technology costs refer to Chapter 3 (section 3.3.2.2)
of the joint TSD. EPA fuel cost estimates have also been updated using the recent AEO 2010
6-1
-------
Regulatory Impact Analysis
Early Release. For more information about the updated fuel prices refer to Chapter 4 of the
joint TSD. EPA has not changed technology cost or fuel cost estimates in response to
comment and, in general, commenters agreed with or otherwise did not question EPA's cost
estimates and methodology. This is true not only for our estimates of 2016 MY costs but also
our estimate of costs for the intermediate years 2012-2015.
6.1 Vehicle Program Costs
Chapter 4 of this RIA presents the outputs of the OMEGA model for the model year
2016. Here, EPA builds on those results and calculates estimated costs for each model year
beginning with 2012 and going through 2050. This is done on a per-vehicle basis and an
annual basis. Costs here include costs associated with the A/C credit program. For details on
the individual technology costs refer to Chapter 3 of the joint TSD. For details on the
OMEGA model inputs (i.e., how the individual technology costs are combined into package
costs) refer to Chapter 1 of this RIA. For details on the A/C costs, refer to Chapter 2 of this
RIA.
6.1.1 Vehicle Compliance Costs on a Per-Vehicle Basis
As stated above, Chapter 4 of this RIA presents the estimated cost per 2016 MY
vehicle for each manufacturer. Those 2016 MY costs are reproduced in Table 6-1. To
estimate the cost per vehicle for model years 2012 through 2015, EPA projected CO2
levels for each manufacturer's fleet for each model year 2011 though 2016. Those
COi levels are presented in
Table 6-2 for cars and Table 6-3 for trucks.A
A Note that the 2012-2015 CO2 levels are estimates based upon assumptions of manufacturer fleetwide CO2
averages in 2011, which are extrapolated from the 2008 baseline fleet. Consequently, the average CO2 emission
levels for some manufacturers are potentially too high for the 2011MY 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 Subaru as shown in Table 6-5 is approximately 45% 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.
6-2
-------
Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-1 Cost per Car and Truck, including A/C, for the 2016 MY Relative to the Cost of Complying
with the 2011 CAFE Standards (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
$/CAR
$1,557
$1,128
$1,535
$1,108
$898
$634
$802
$667
$854
$817
$686
$1,506
$961
$1,014
$1,180
$380
$1,847
$869
$/TRUCK
$1,195
$1,501
$930
$1,441
$1,581
$473
$425
$247
$537
$1,217
$1,118
$758
$789
$536
$679
$609
$971
$1,098
Table 6-2 Projected CO2 Levels for MYs 2011-2016, Cars Only (g/mi CO2)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2011MY
313.8
314.7
323.2
311.3
305.4
265.0
273.8
280.6
290.2
286.8
280.0
339.0
286.4
285.2
348.2
253.1
291.5
288.0
2012MY
297.0
288.4
306.5
276.9
278.2
247.4
255.7
257.2
261.6
262.1
263.1
318.7
259.8
266.3
328.9
243.7
276.6
266.8
2013MY
282.1
271.6
291.7
268.8
270.4
236.6
244.5
244.3
252.9
252.7
255.8
300.3
246.8
249.3
311.6
236.3
263.6
256.0
2014MY
266.3
253.9
276.0
259.9
261.5
224.7
232.3
230.4
243.9
239.6
247.5
281.0
238.6
231.4
293.4
227.8
249.6
245.2
2015MY
250.4
240.0
260.2
247.1
248.3
215.2
222.5
219.4
232.2
226.6
236.3
261.6
230.4
213.4
275.1
219.4
235.6
233.7
2016MY
236.3
227.9
246.3
233.4
241.3
207.6
214.5
213.1
218.2
223.3
223.2
244.1
218.2
197.3
258.6
212.8
223.5
223.8
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
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2011MY
369.8
380.9
423.5
394.5
402.5
354.9
356.8
366.9
341.2
336.9
399.3
364.2
335.8
338.3
377.5
375.4
420.3
384.4
2012MY
350.1
355.7
396.9
373.3
384.1
338.7
344.1
356.1
324.7
311.6
361.0
356.3
307.5
332.7
365.2
360.6
400.1
364.7
2013MY
332.6
339.9
372.4
362.5
372.5
328.8
336.6
350.4
312.5
300.8
350.8
350.5
293.6
329.2
355.1
347.9
382.0
352.8
2014MY
313.9
322.9
346.9
350.8
362.2
317.8
327.9
343.7
302.3
292.0
339.5
343.7
284.8
324.7
343.9
334.1
362.9
339.7
2015MY
295.3
309.5
321.3
332.1
344.0
308.9
320.8
337.9
286.5
277.0
321.0
336.8
275.9
320.2
332.8
320.3
343.7
324.5
2016MY
278.7
298.2
297.8
298.3
305.1
302.0
315.6
335.2
285.6
264.0
299.8
332.0
263.0
317.7
323.6
308.6
326.6
302.5
The achieved CCh levels for 2012-2015 were derived using a similar process to that
described in Chapter 5 of this RIA. As in Chapter 5, EPA estimated model year specific
emission targets based on the GHG standard curves and each manufacturer's projected fleet
mix. From these targets, EPA adjusted CC>2 emissions by the impact of anticipated FFV
credit, A/C credit, and TLAAS usage. For the emission analysis presented in RIA Chapter 5,
EPA also estimated the impact of credit transfers which increased net emissions; i.e., several
manufacturers 2008 baseline vehicles overcomplied with the 2011 MY CAFE standard for
cars in the early program years. The credits associated with this overcompliance with the car
standards could be traded to these manufacturers' trucks fleets with the result of a net
emission increase relative to a scenario that restricted transfers (i.e., required each
manufacturer's trucks to meet the 2011 CAFE standard while retaining over-compliance for
cars). Credit transfers beyond baseline overcompliance is environmentally neutral and were
not required in the emission analysis presented in Chapter 5.
For the cost analysis presented here, EPA additionally considered environmentally
neutral credit transfers. In order to reduce overall compliance costs, manufacturers may
choose to overcomply with the GHG standard in either car or truck categories and "trade" (on
a VMT weighted basis), their overcompliance to the other category. As detailed in RIA
Chapter 4, OMEGA incorporates this flexibility and projects technology application based on
the most cost effective path to compliance. Thus, the fleetwide COi achieved emission levels
from RIA chapter 5 had to be adjusted to incorporate environmentally neutral credit transders.
In these adjustments, the VMT adjusted fleet average CCh by manufacturer remained the
same, but the average emission levels for cars and trucks may be above or below the
applicable CO2 standards and was based on the results of the OMEGA modeling.
6-4
-------
Vehicle Program Costs Including Fuel Consumption Impacts
The cost effective achieved levels for the intermediate years were derived in the
following manner. Car and truck COi emissions in MY 2011 were taken directly from the
reference fleet file.3 This fleet consists of 2008 MY vehicles, with sales projected to 2011 and
with emissions reduced to the extent necessary for each manufacturer to comply with the
2011 MY CAFEE standards, with one exception. This exception was that those
manufacturers which traditionally have paid CAFE fines in lieu of compliance were allowed
to do so when NHTSA's VOLPE model estimated that it was less expensive for these
manufacturers to pay fines instead of adding additional technology to meet the 2011 MY
standards. MY 2016 achieved CO2 was determined from the OMEGA output described in
RIA Chapter 4. To determine the COi emissions by manufacturer for the intermediate years,
an interpolation was performed between these two points. Two different forms of
interpolation were used, as appropriate. Generally, for manufacturers projected by the
OMEGA modeling to achieve the 2016 MY standards, the change between each
manufacturer's 2011 and 2016 emission levels was weighted by the percent change between
their fleet average (i.e., car plus truck) standard for each year (as determined in RIA Chapter
5) relative to their 2011 MY emission level. For manufacturers that are projected by the
OMEGA modeling to not achieve their MY 2016 standard, we assumed a linear improvement
between their 2011 MY and 2016 MY emission levels (i.e., 20% of the total change between
2011-2016 emission levels was applied each year).
Several manufacturers, including Subaru, Kia, Mazda, 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.4
We then used these COi values to generate ratios that could be applied to the 2016
MY costs to arrive at cost estimates for each of the intervening years. However, it is
important to remember that the technology costs and, subsequently, the package costs in the
2016 MY have undergone some adjustments to account for learning effects as described in
Chapter 3 of the joint TSD. EPA 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. This was
also done for the years following 2016 to reflect the effects of the near term and long term
indirect cost multipliers (ICMs) as described in Chapter 3 of the 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 $898 in the 2016 MY while achieving a COi
average of 241 g/mi; for the 2011 and 2012 MYs, GM cars are projected to achieve a CO2
average of 305 and 278, respectively. The ratio (305-278)7(305-241) can be applied to GM's
2016 cost of $898, and then apply the 2012 relative to 2016 cost factor of 119%, to arrive at
an estimated 2012 cost of $453.c This process is carried out 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
B
Ibid.
c Numbers in the text are rounded for clarity so results using numbers shown in the text may not match those in
tables.
6-5
-------
Regulatory Impact Analysis
for cars and trucks combined. Table 6-8 shows the estimated industry average cost per car,
cost per truck, and cost per vehicle (car/truck combined) for the 2012 and later model years.0
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 2016
119%
117%
109%
102%
100%
100%
100%
100%
100%
100%
94%
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. Higher new vehicle costs are included for these latter items but do not
account for higher replacement costs during vehicle lifetimes even though oil is changed many times and tires
are changed once or twice. Using the incremental increase in maintenance costs (and savings) and discounting
them back to present value would have little impact on the present value of the costs presented here. Note also
that the expected penetration of A/C control technology is 28% in 2012 but was 25% in our emission modeling
work (similarly small discrepancies exist for 2013-2015). The slightly lower penetration number used in the
emission modeling indicates a slight underestimation of the emission reductions from MY 2012, and
consequently a slight underestimation of the costs in 2012-2015.
6-6
-------
Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-5 Cost per Car, including A/C, by Manufacturer Relative to the Cost of Complying with the 2011
CAFE Standards (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2012MY
$402
$408
$397
$583
$453
$230
$291
$275
$404
$378
$243
$383
$445
$259
$302
$105
$482
$342
2013MY
$746
$656
$735
$708
$575
$367
$464
$420
$518
$514
$342
$718
$653
$484
$563
$186
$887
$507
2014MY
$1,042
$862
$1,027
$797
$670
$485
$612
$540
$599
$662
$428
$1,004
$734
$677
$787
$260
$1,240
$631
2015MY
$1,300
$991
$1,282
$932
$816
$560
$708
$616
$702
$790
$538
$1,252
$805
$844
$982
$324
$1,547
$749
2016MY
$1,557
$1,128
$1,535
$1,108
$898
$634
$802
$667
$854
$817
$686
$1,506
$961
$1,014
$1,180
$380
$1,847
$869
Table 6-6 Cost per Truck, including A/C, by Manufacturer Relative to the Cost of Complying with the
2011 CAFE Standards (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2012MY
$307
$543
$235
$377
$355
$172
$156
$101
$190
$502
$512
$222
$365
$174
$184
$161
$249
$314
2013MY
$571
$871
$442
$561
$569
$273
$244
$150
$324
$706
$638
$377
$535
$276
$330
$294
$464
$496
2014MY
$799
$1,146
$618
$714
$713
$361
$325
$197
$410
$818
$732
$527
$603
$385
$461
$410
$649
$652
2015MY
$996
$1,321
$771
$953
$968
$419
$379
$230
$538
$1,021
$898
$657
$662
$481
$575
$512
$810
$820
2016MY
$1,195
$1,501
$930
$1,441
$1,581
$473
$425
$247
$537
$1,217
$1,118
$758
$789
$536
$679
$609
$971
$1,098
6-7
-------
Regulatory Impact Analysis
Table 6-7 Cost per Vehicle (car/truck combined), including A/C, by Manufacturer Relative to the Cost of
Complying with the 2011 CAFE Standards (2007 dollars)
MANUFACTURER
BMW
Chrysler
Daimler
Ford
General Motors
Honda
Hyundai
Kia
Mazda
Mitsubishi
Nissan
Porsche
Subaru
Suzuki
Tata
Toyota
Volkswagen
Overall
2012MY
$363
$493
$337
$511
$406
$206
$273
$197
$356
$400
$334
$334
$419
$239
$295
$125
$437
$331
2013MY
$679
$781
$622
$658
$572
$330
$419
$312
$471
$545
$449
$614
$615
$432
$524
$222
$794
$503
2014MY
$959
$1,020
$871
$768
$691
$439
$553
$386
$559
$685
$537
$825
$692
$583
$728
$312
$1,106
$639
2015MY
$1,209
$1,171
$1,089
$940
$889
$507
$656
$459
$672
$826
$657
$1,009
$762
$719
$913
$387
$1,412
$774
2016MY
$1,453
$1,328
$1,312
$1,228
$1,219
$574
$745
$501
$799
$876
$823
$1,206
$912
$855
$1,099
$455
$1,693
$948
6-8
-------
Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-8 Industry Average Cost per Car, Truck, and Combined by Year Relative to the Cost of
Complying with the 2011 CAFE Standards (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
$/CAR
$342
$507
$631
$749
$869
$869
$869
$869
$869
$869
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$817
$/TRUCK
$314
$496
$652
$820
$1,098
$1,098
$1,098
$1,098
$1,098
$1,098
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$1,032
$/VEHICLE
$331
$503
$639
$774
$948
$947
$945
$943
$940
$939
$882
$881
$881
$880
$880
$879
$879
$878
$878
$877
$877
$876
$876
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
$875
6-9
-------
Regulatory Impact Analysis
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, annual costs can be calculated 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
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
289,631
409,462
211,652
1,468,182
1,586,094
906,096
376,284
299,611
283,128
110,284
824,030
41,117
183,486
72,297
36,377
1,591,054
434,412
9,123,197
2013MY
293,905
426,454
202,559
1,485,801
1,544,975
1,064,848
395,573
326,652
329,911
104,555
831,607
43,299
175,170
81,781
50,527
1,941,480
498,641
9,797,738
2014MY
369,979
411,319
244,554
1,567,762
1,452,559
1,087,076
395,515
427,191
378,291
88,150
854,131
34,024
184,521
90,597
49,316
2,079,011
517,978
10,231,974
2015MY
411,653
392,483
263,751
1,542,470
1,487,318
912,434
511,236
538,717
413,328
82,310
925,478
32,426
204,746
100,600
63,751
2,176,644
567,711
10,627,055
2016MY
422,874
399,762
270,940
1,559,310
1,514,479
930,350
518,445
548,055
420,516
82,688
946,518
33,309
206,903
103,003
65,489
2,226,522
583,185
10,832,348
E Note that the vehicle compliance costs presented here do not include costs associated with upgrading testing
facilities to accommodate N2O testing. Including those costs would have very little impact on the costs
presented here for new vehicle technology.
6-10
-------
Vehicle Program Costs Including Fuel Consumption Impacts
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
204,197
692,115
108,053
851,877
1,510,917
634,705
58,164
87,643
60,783
48,290
405,017
13,190
97,935
4,593
29,647
886,621
104,842
5,798,588
2013MY
183,550
594,092
114,531
940,080
1,536,070
676,729
101,529
102,773
64,784
46,179
391,572
14,608
89,944
16,557
33,749
990,315
141,421
6,038,484
2014MY
191,010
514,802
136,455
965,589
1,336,797
634,606
103,857
114,423
67,780
52,835
406,045
16,033
99,293
20,060
40,294
1,095,949
151,992
5,947,819
2015MY
175,612
475,312
129,878
936,781
1,379,813
560,745
94,606
118,391
74,213
56,896
391,733
17,145
116,055
20,547
43,703
1,107,261
127,888
5,826,579
2016MY
170,749
462,150
126,281
910,840
1,341,604
545,217
91,986
115,113
72,158
55,320
380,886
16,670
117,295
19,978
42,493
1,076,598
124,346
5,669,683
Table 6-11 Estimated Annual Costs by Manufacturer, including A/C, for Cars Relative to the Cost of
Complying with the 2011 CAFE Standards ($Millions 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
$170
$80
$860
$720
$210
$110
$80
$110
$40
$200
$20
$80
$20
$10
$170
$210
$3,120
2013MY
$220
$280
$150
$1,050
$890
$390
$180
$140
$170
$50
$280
$30
$110
$40
$30
$360
$440
$4,970
2014MY
$390
$350
$250
$1,250
$970
$530
$240
$230
$230
$60
$370
$30
$140
$60
$40
$540
$640
$6,460
2015MY
$540
$390
$340
$1,440
$1,210
$510
$360
$330
$290
$70
$500
$40
$160
$80
$60
$710
$880
$7,960
2016MY
$660
$450
$420
$1,730
$1,360
$590
$420
$370
$360
$70
$650
$50
$200
$100
$80
$850
$1,080
$9,410
6-11
-------
Regulatory Impact Analysis
Table 6-12 Estimated Annual Costs by Manufacturer, including A/C, for Trucks Relative to the Cost of
Complying with the 2011 CAFE Standards ($Millions 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
$380
$30
$320
$540
$110
$10
$10
$10
$20
$210
$0
$40
$0
$10
$140
$30
$1,820
2013MY
$100
$520
$50
$530
$870
$180
$20
$20
$20
$30
$250
$10
$50
$0
$10
$290
$70
$2,990
2014MY
$150
$590
$80
$690
$950
$230
$30
$20
$30
$40
$300
$10
$60
$10
$20
$450
$100
$3,880
2015MY
$170
$630
$100
$890
$1,340
$240
$40
$30
$40
$60
$350
$10
$80
$10
$30
$570
$100
$4,780
2016MY
$200
$690
$120
$1,310
$2,120
$260
$40
$30
$40
$70
$430
$10
$90
$10
$30
$660
$120
$6,230
Table 6-13 Estimated Annual Costs by Manufacturer, including A/C, for Cars and Trucks Combined
Relative to the Cost of Complying with the 2011 CAFE Standards ($Millions 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
$550
$110
$1,180
$1,260
$320
$120
$90
$120
$60
$410
$20
$120
$20
$20
$310
$240
$4,940
2013MY
$320
$800
$200
$1,580
$1,760
$570
$200
$160
$190
$80
$530
$40
$160
$40
$40
$650
$510
$7,960
2014MY
$540
$940
$330
$1,940
$1,920
$760
$270
$250
$260
$100
$670
$40
$200
$70
$60
$990
$740
$10,340
2015MY
$710
$1,020
$440
$2,330
$2,550
$750
$400
$360
$330
$130
$850
$50
$240
$90
$90
$1,280
$980
$12,740
2016MY
$860
$1,140
$540
$3,040
$3,480
$850
$460
$400
$400
$140
$1,080
$60
$290
$110
$110
$1,510
$1,200
$15,640
6-12
-------
Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-14 Annual Sales & Costs for Cars & Trucks Relative to the Cost of Complying with the 2011
CAFE Standards (Monetary Values in 2007 dollars)
YEAR
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV, 3%
NPV, 7%
CAR SALES
9,123,197
9,797,738
10,231,974
10,627,055
10,832,348
10,694,686
10,688,658
10,930,973
11,387,037
11,411,597
11,406,245
11,512,039
11,744,448
11,997,261
12,196,567
12,379,457
12,554,527
12,711,837
12,888,819
13,022,913
13,193,076
13,386,235
13,601,889
13,814,706
13,937,501
14,061,386
14,186,373
14,312,471
14,439,690
14,568,040
14,697,530
14,828,171
14,959,974
15,092,948
15,227,104
15,362,453
15,499,005
15,636,770
15,775,760
TRUCK SALES
5,798,588
6,038,484
5,947,819
5,826,579
5,669,683
5,490,258
5,281,918
5,191,411
5,154,530
5,048,217
4,938,711
4,900,413
4,938,251
4,968,893
4,995,378
5,018,975
5,025,873
5,027,891
5,068,246
5,068,435
5,078,420
5,109,052
5,135,430
5,151,342
5,197,131
5,243,327
5,289,933
5,336,953
5,384,392
5,432,252
5,480,537
5,529,252
5,578,399
5,627,984
5,678,009
5,728,479
5,779,398
5,830,769
5,882,596
CAR COSTS
($MILLIONS)
$3,120
$4,970
$6,460
$7,960
$9,410
$9,290
$9,290
$9,500
$9,900
$9,920
$9,320
$9,400
$9,590
$9,800
$9,960
$10,110
$10,260
$10,380
$10,530
$10,640
$10,780
$10,940
$11,110
$11,290
$11,390
$11,490
$11,590
$11,690
$11,800
$11,900
$12,010
$12,110
$12,220
$12,330
$12,440
$12,550
$12,660
$12,770
$12,890
$226,730
$124,010
TRUCK COSTS
($MILLIONS)
$1,820
$2,990
$3,880
$4,780
$6,230
$6,030
$5,800
$5,700
$5,660
$5,540
$5,100
$5,060
$5,100
$5,130
$5,160
$5,180
$5,190
$5,190
$5,230
$5,230
$5,240
$5,270
$5,300
$5,320
$5,370
$5,410
$5,460
$5,510
$5,560
$5,610
$5,660
$5,710
$5,760
$5,810
$5,860
$5,910
$5,970
$6,020
$6,070
$119,200
$67,850
TOTAL COSTS
($MILLIONS)
$4,940
$7,960
$10,340
$12,740
$15,640
$15,320
$15,090
$15,200
$15,560
$15,460
$14,420
$14,460
$14,690
$14,930
$15,120
$15,290
$15,450
$15,570
$15,760
$15,870
$16,020
$16,210
$16,410
$16,610
$16,760
$16,900
$17,050
$17,200
$17,360
$17,510
$17,670
$17,820
$17,980
$18,140
$18,300
$18,460
$18,630
$18,790
$18,960
$345,940
$191,860
6-13
-------
Regulatory Impact Analysis
6.2 Cost per Ton of Emissions Reduced
We have calculated the cost per ton of GHG (COi equivalent, or CC^e) reductions
associated with this GHG rule using the costs shown in Table 6-14 and the emissions
reductions described in Chapter 5. The cost per metric ton of GHG emissions reductions in
the years 2020, 2030, 2040, and 2050 is calculated using the annual vehicle compliance costs
and emission reductions for each of those years. The value in 2050 represents the long-term
cost per ton of the emissions reduced. EPA has also calculated the cost per metric ton of
GHG emission reductions including the savings associated with reduced fuel consumption
(see section 6.3 below). This latter calculation does not include the other benefits associated
with this rule such as those associated with criteria pollutant reductions or energy security
benefits as discussed in Chapter 8 of this RIA. By including the fuel savings in the cost
estimates, the cost per ton is less than $0, since the estimated value of fuel savings outweighs
the vehicle program costs. With regard to the CH4 and NiO standards, since these standards
would be emissions caps designed to ensure that manufacturers do not backslide from current
levels, the costs associated with the standards were not estimated (since the standards would
not require any change from current practices nor is it estimated they would result in
emissions reductions ).
The results for CC^e costs per ton under the final rule are shown in Table 6-15.
Table 6-15 Annual Cost Per Metric Ton of CO2e Reduced, in $2007 dollars
Year
2020
2030
2040
2050
Vehicle
Compliance
Costa
($Millions)
$15,600
$15,800
$17,400
$19,000
Fuel Savings b
($Millions)
-$35,700
-$79,800
-$119,300
-$171,200
CCh-equivalent
Reduction
(Million metric tons)
160
310
400
510
Cost per Ton -
Vehicle
Program only
$100
$50
$40
$40
Cost per Ton -
Vehicle Program
with Fuel Savings
-$130
-$210
-$250
-$300
a Costs here include vehicle compliance costs and do not include any fuel savings
bFuel savings calculated using pre-tax fuel prices.
6.3 Fuel Consumption Impacts
In this section, EPA presents the impact of the final rule 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
rule.
The new COi standards will result in significant improvements in the fuel efficiency of
affected vehicles. Drivers of those vehicles will see corresponding savings associated with
Including those costs would have very little impact on the costs presented here for new vehicle technology.
6-14
-------
Vehicle Program Costs Including Fuel Consumption Impacts
reduced expenditures for fuel. EPA has estimated the impacts on fuel consumption for both
the tailpipe CC>2 standards and the A/C credit program. To do this, fuel consumption is
calculated using both current CO2 emission levels and the new CO2 standards. The difference
between these estimates represents the net savings from the new COi standards.
The expected impacts on fuel consumption are shown in Table 6-16. The gallons
shown in the tables reflect impacts from the new CC>2 standards, including the A/C credit
program, and include increased consumption resulting from the rebound effect. Using these
fuel consumption estimates, the monetized fuel savings associated with the new CC>2
standards can be calculated. To do this, the reduced fuel consumption in each year is
multiplied by the corresponding estimated average fuel price in that year, using the reference
case taken from the AEO 2010 Early Release. AEO is the government consensus estimate
used by NHTSA and many other government agencies to estimate the projected price of fuel.
The calculation has been done using both the pre-tax and the post-tax fuel prices. The latter
of these is what consumers actually pay for the fuel and, therefore, the post-tax fuel savings
are those savings that consumers will realize. The pre-tax fuel savings represent the savings
to society.
Table 6-16 Annual Fuel Consumption Impacts of the 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
GALLONS
(MILLIONS)
500
1,300
2,300
3,800
5,700
7,500
9,200
10,900
12,600
14,200
15,600
17,100
18,400
19,700
20,800
22,000
22,900
23,900
24,700
25,600
26,400
FUEL PRICE
EXCLUDING
TAXES
($/GALLON)
$2.08
$2.21
$2.45
$2.56
$2.61
$2.68
$2.75
$2.80
$2.84
$2.89
$2.92
$2.96
$2.99
$3.01
$3.05
$3.09
$3.13
$3.18
$3.23
$3.25
$3.29
FUEL PRICE
INCLUDING
TAXES
($/GALLON)
$2.61
$2.84
$2.95
$3.00
$3.07
$3.13
$3.19
$3.22
$3.27
$3.29
$3.34
$3.37
$3.38
$3.42
$3.46
$3.49
$3.54
$3.59
$3.60
$3.64
$3.69
PRE-TAX
FUEL
SAVINGS
($MILLIONS)
$1,100
$2,900
$5,700
$9,600
$14,800
$20,100
$25,400
$30,600
$35,700
$40,900
$45,600
$50,500
$55,100
$59,200
$63,600
$67,900
$71,800
$76,000
$79,800
$83,100
$86,800
POST-TAX
FUEL SAVINGS
($MILLIONS)
$1,400
$3,800
$6,900
$11,300
$17,400
$23,500
$29,400
$35,200
$41,100
$46,700
$52,200
$57,400
$62,200
$67,300
$72,100
$76,600
$81,200
$85,600
$89,100
$93,200
$97,300
6-15
-------
Regulatory Impact Analysis
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
NPV,
3%
NPV,
7%
27,200
27,900
28,700
29,500
30,300
31,000
31,800
32,600
33,400
34,300
35,100
36,000
36,800
37,700
38,700
39,600
40,500
41,500
$3.34
$3.37
$3.42
$3.48
$3.53
$3.57
$3.61
$3.66
$3.70
$3.75
$3.79
$3.84
$3.89
$3.93
$3.98
$4.03
$4.08
$4.12
$3.72
$3.77
$3.83
$3.87
$3.91
$3.95
$3.99
$4.04
$4.08
$4.12
$4.17
$4.21
$4.26
$4.30
$4.35
$4.39
$4.44
$4.49
$90,700
$94,200
$98,200
$102,700
$106,700
$110,800
$115,000
$119,300
$123,800
$128,400
$133,200
$138,100
$143,200
$148,400
$153,900
$159,500
$165,300
$171,200
$1,545,600
$672,600
$101,100
$105,200
$109,900
$114,100
$118,400
$122,700
$127,100
$131,700
$136,400
$141,200
$146,300
$151,400
$156,800
$162,300
$168,100
$173,900
$180,000
$186,300
$1,723,900
$755,700
As shown in Table 6-16, we are projecting that consumers will realize very large fuel
savings as a result of the new COi standards. There are several ways to view this value.
Some, as demonstrated below in Chapter 8 of this RIA, view these fuel savings as a reduction
in the cost of owning a vehicle, whose full benefits consumers realize. This approach
assumes that, regardless of how consumers in fact make their decisions on how much fuel
economy to purchase, they will necessarily gain these fuel savings. Another view says that
consumers do not necessarily value fuel savings as equal to the results of this calculation,
notwithstanding actual dollars accruing to them. Instead, consumers may either undervalue or
overvalue fuel economy relative to these savings, based on their personal preferences. This
issue is discussed further in Section 8.1.2 of this RIA.
If the analysis is limited to the five model years 2012-2016—in other words, to the
fuel consumption savings during the vehicle lifetimes of those five model years, the results
would be as shown in Table 6-17.
6-16
-------
Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-17 Annual Fuel Savings for 2012-2016 MY Vehicles Using Pre-tax Fuel Prices ($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%
2012MY
$1,300
$1,300
$1,400
$1,500
$1,400
$1,400
$1,300
$1,300
$1,200
$1,100
$1,000
$900
$800
$700
$500
$400
$400
$300
$200
$200
$200
$100
$100
$100
$100
$100
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$15,600
$12,100
2013MY
$2,000
$2,200
$2,200
$2,100
$2,100
$2,000
$2,000
$1,800
$1,700
$1,600
$1,500
$1,300
$1,100
$1,000
$800
$700
$500
$400
$400
$300
$200
$200
$200
$100
$100
$100
$100
$100
$100
$100
$0
$0
$0
$0
$0
$0
$0
$0
$23,300
$18,100
2014MY
$2,900
$3,000
$2,900
$2,900
$2,800
$2,700
$2,600
$2,500
$2,300
$2,200
$2,000
$1,800
$1,500
$1,300
$1,100
$900
$700
$600
$500
$400
$300
$300
$200
$200
$200
$100
$100
$100
$100
$100
$100
$100
$0
$0
$0
$0
$0
$31,600
$24,600
2015MY
$4,300
$4,300
$4,200
$4,200
$4,100
$3,900
$3,800
$3,500
$3,300
$3,100
$2,800
$2,500
$2,200
$1,800
$1,500
$1,200
$1,000
$800
$700
$600
$500
$400
$300
$300
$200
$200
$100
$100
$100
$100
$100
$100
$100
$100
$100
$0
$45,300
$35,400
2016MY
$6,000
$5,900
$5,900
$5,800
$5,600
$5,400
$5,100
$4,900
$4,600
$4,200
$3,900
$3,500
$3,000
$2,500
$2,100
$1,700
$1,400
$1,100
$900
$800
$600
$500
$400
$400
$300
$300
$200
$200
$200
$100
$100
$100
$100
$100
$100
$62,500
$48,800
SUM
$1,300
$3,400
$6,500
$10,900
$16,700
$16,600
$16,300
$15,800
$15,200
$14,500
$13,600
$12,800
$11,800
$10,600
$9,400
$8,200
$6,900
$5,800
$4,700
$3,900
$3,200
$2,600
$2,100
$1,700
$1,500
$1,200
$1,000
$900
$700
$600
$500
$400
$400
$300
$300
$300
$200
$200
$100
$178,300
$139,000
6-17
-------
Regulatory Impact Analysis
6.4 Vehicle Program Cost Summary
The vehicle program costs consist of the vehicle compliance costs relative to the cost
of complying with the 2011 CAFE standards, and the fuel savings that would result from the
reduction in fuel consumption. These costs are summarized in Table 6-18.
6-18
-------
Vehicle Program Costs Including Fuel Consumption Impacts
Table 6-18 Annual Vehicle Program Costs Including Fuel Savings Using Post-Tax Fuel Prices ($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%
VEHICLE
COMPLIANCE
COSTS
$4,900
$8,000
$10,300
$12,700
$15,600
$15,300
$15,100
$15,200
$15,600
$15,500
$14,400
$14,500
$14,700
$14,900
$15,100
$15,300
$15,500
$15,600
$15,800
$15,900
$16,000
$16,200
$16,400
$16,600
$16,800
$16,900
$17,100
$17,200
$17,400
$17,500
$17,700
$17,800
$18,000
$18,100
$18,300
$18,500
$18,600
$18,800
$19,000
$345,900
$191,900
FUEL SAVINGS)
-$1,400
-$3,800
-$6,900
-$11,300
-$17,400
-$23,500
-$29,400
-$35,200
-$41,100
-$46,700
-$52,200
-$57,400
-$62,200
-$67,300
-$72,100
-$76,600
-$81,200
-$85,600
-$89,100
-$93,200
-$97,300
-$101,100
-$105,200
-$109,900
-$114,100
-$118,400
-$122,700
-$127,100
-$131,700
-$136,400
-$141,200
-$146,300
-$151,400
-$156,800
-$162,300
-$168,100
-$173,900
-$180,000
-$186,300
-$1,723,900
-$755,700
TOTAL
$3,500
$4,200
$3,400
$1,400
-$1,800
-$8,200
-$14,300
-$20,000
-$25,500
-$31,200
-$37,800
-$42,900
-$47,500
-$52,400
-$57,000
-$61,300
-$65,700
-$70,000
-$73,300
-$77,300
-$81,300
-$84,900
-$88,800
-$93,300
-$97,300
-$101,500
-$105,600
-$109,900
-$114,300
-$118,900
-$123,500
-$128,500
-$133,400
-$138,700
-$144,000
-$149,600
-$155,300
-$161,200
-$167,300
-$1,378,000
-$563,800
6-19
-------
Regulatory Impact Analysis
References
All references can be found in the EPA DOCKET: EPA-HQ-OAR-2009-0472.
1 EPA-420-R-09-020, EPA docket number EPA-HQ-OAR-2009-0472-11282; peer review
report dated November 6, 2009, is at EPA-HQ-OAR-2009-0472-11285.
2 "Binning of FEV Costs to GDI, Turbo-charging, and Engine Downsizing," memorandum to
Docket EPA-HQ-OAR-2009-0472, from Michael Olechiw, U.S. EPA, dated March 25, 2010.
3 U.S. EPA. Baseline and Reference Fleet File, as documented in TSD chapter 1. August
2009.
4 US EPA 2010. Cost effective achieved levels spreadsheet.
6-20
-------
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: particulate matter, ozone, nitrogen oxides (NOx), sulfur oxides (SOx), carbon
monoxide and air toxics. These pollutants would not be directly regulated by the standards,
but the standards would affect emissions of these pollutants and precursors.
7.1.1.1 Particulate Matter
7.1.1.1.1 Background
Particulate 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 jam), 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 jam and less than or equal to 10 jam, or PMio-2.5).
Ultrafine particles (UFPs) are a subset of fine particles, generally less than 100 nanometers
(0.1 ^im) in aerodynamic diameter.
Particles span many sizes and shapes and consist of numerous 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 PM2.5 may vary greatly with time, region, meteorology and source
category. Thus, PM2.5 may include a complex mixture of different chemicals including
7-1
-------
Regulatory Impact Analysis
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
7.1.1.1.2 Paniculate Matter Health Effects
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 information
and conclusions in the Integrated Science Assessment (ISA) for Particulate Matter (December
2009) prepared by EPA's Office of Research and Development (ORD).B
The ISA concludes that ambient concentrations of PM are associated with a number of
adverse health effects.c The ISA characterizes the weight of evidence for different health
effects associated with three PM size ranges: PM2.5, PMio-2.5, and UFPs. The discussion
below highlights the ISA's conclusions pertaining to these three size fractions of PM,
considering variations in both short-term and long-term exposure periods.
7.1.1.1.2.1 Effects Associated with PM2.5
Short-term Exposure
The ISA concludes that cardiovascular effects and all-cause cardiovascular- and
respiratory-related mortality are causally associated with short-term exposure to PM2.5.2 It
also concludes that respiratory effects are likely to be causally associated with short-term
exposure to PMi.5, including respiratory emergency department (ED) visits and hospital
admissions for chronic obstructive pulmonary disease (COPD), respiratory infections, and
asthma; and exacerbation of respiratory symptoms in asthmatic children.
Long-term Exposure
The ISA concludes that there are causal associations between long-term exposure to
PM2.5 and cardiovascular effects, such as the development/progression of cardiovascular
disease (CVD), and premature mortality, particularly from cardiopulmonary causes.3 It also
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 ISA is available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=216546
c The ISA evaluates the health evidence associated with different health effects, assigning one of five "weight of
evidence" determination: causal relationship, likely to be a causal relationship, suggestive of a causal
relationship, inadequate to infer a causal relationship, and not likely to be a causal relationship. For definitions
of these levels of evidence, please refer to Section 1.5 of the ISA. The following text summarizes only those
health effects with at least a "suggestive" weight of evidence determination.
7-2
-------
Environmental and Health Impacts
concludes that long-term exposure to PM2.5 is likely to be causally associated with respiratory
effects, such as reduced lung function growth, increased respiratory symptoms, and asthma
development. The ISA characterizes the evidence as suggestive of a causal relationship for
associations between long-term PMi.5 exposure and reproductive and developmental
outcomes, such as low birth weight and infant mortality. It also characterizes the evidence as
suggestive of a causal relationship between PMi.5 and cancer incidence, mutagenicity, and
genotoxicity.
7.1.1.1.2.2 Effects Associated with PM10.2.5
The ISA summarizes evidence related to short-term exposure to PMio-2.5- PMio-2.5 is
the fraction of PMi0 particles that is larger than PM2.5.4 The ISA concludes that available
evidence is suggestive of a causal relationship between short-term exposures to PMio-2.5 and
cardiovascular effects, such as hospitalizations for ischemic heart disease. It also concludes
that the available evidence is suggestive of a causal relationship between short-term exposures
to PMio-2.5 and respiratory effects, including respiratory-related ED visits and hospitalizations
and pulmonary inflammation. The ISA also concludes that the available literature suggests a
causal relationship between short-term exposures to PMi0-2.5 and mortality. Data are
inadequate to draw conclusions regarding health effects associated with long-term exposure to
PMiO-2.5-5
7.1.1.1.2.3 Effects Associated with Ultrafme Particles
The ISA concludes that the evidence is suggestive of a causal relationship between
short-term exposures to UFPs and cardiovascular effects, including changes in heart rhythm
and vasomotor function (the ability of blood vessels to expand and contract).6
The ISA also concludes that there is suggestive evidence of a causal relationship
between short-term UFP exposure and respiratory effects. The types of respiratory effects
examined in epidemiologic studies include respiratory symptoms and asthma hospital
admissions, the results of which are not entirely consistent. There is evidence from
toxicological and controlled human exposure studies that exposure to UFPs may increase lung
inflammation and produce small asymptomatic changes in lung function. Data are inadequate
to draw conclusions regarding health effects associated with long-term exposure to UFPs.7
7.1.1.2 Ozone
7.1.1.2.1 Background
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.
7-3
-------
Regulatory Impact Analysis
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 (NC^); 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.8'9 We are relying on the data
and conclusions in the ozone AQCD and staff paper, regarding the health effects associated
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
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.10 People who appear to be more
susceptible to effects associated with exposure to ozone 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.11'12'13'14'15'16
Repeated exposure to ozone can increase susceptibility to respiratory infection and lung
inflammation and can aggravate preexisting respiratory diseases, such as asthma.17'18' '20'21
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.22'23'24'25
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.26 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.27 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.28'29'30'31'32'33'34'35 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.36'37'38'39
7-5
-------
Regulatory Impact Analysis
7.1.1.3 Nitrogen Oxides and Sulfur Oxides
7.1.1.3.1 Background
Sulfur dioxide (862), a member of the sulfur oxide (SOx) family of gases, is formed
from burning fuels containing sulfur (e.g., coal or oil), extracting gasoline from oil, or
extracting metals from ore. Nitrogen dioxide (NOi) is a member of the nitrogen oxide (NOx)
family of gases. Most NOi is formed in the air through the oxidation of nitric oxide (NO)
emitted when fuel is burned at a high temperature. SOi andNOi 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 ofSOj
This section provides an overview of the health effects associated with SO2.
Additional information on the health effects of SO2 can be found in the EPA Integrated
Science Assessment for Sulfur Oxides.40 Following an extensive evaluation of health
evidence from epidemiologic and laboratory studies, the U.S. EPA has concluded that there is
a causal relationship between respiratory health effects and short-term exposure to SO2. The
immediate effect of SO2 on the respiratory system in humans is bronchoconstriction.
Asthmatics are more sensitive to the effects of SO2 likely resulting from preexisting
inflammation associated with this disease. In laboratory studies involving controlled human
exposures to SO2, respiratory effects have consistently been observed following 5-10 min
exposures at SO2 concentrations > 0.4 ppm in asthmatics engaged in moderate to heavy levels
of exercise, with more limited evidence of respiratory effects among exercising asthmatics
exposed to concentrations as low as 0.2-0.3 ppm. A clear concentration-response relationship
has been demonstrated in these studies following exposures to SO2 at concentrations between
0.2 and 1.0 ppm, both in terms of increasing severity of respiratory symptoms and decrements
in lung function, as well as the percentage of asthmatics adversely affected.
In epidemiologic studies, respiratory effects have been observed in areas where the
mean 24-hour SO2 levels range from 1 to 30 ppb, with maximum 1 to 24-hour average SO2
values ranging from 12 to 75 ppb. Important new multicity studies and several other studies
have found an association between 24-hour average ambient SO2 concentrations and
respiratory symptoms in children, particularly those with asthma. Generally consistent
associations also have been observed between ambient SO2 concentrations and emergency
department visits and hospitalizations for all respiratory causes, particularly among children
and older adults (> 65 years), and for asthma. A limited subset of epidemiologic studies have
examined potential confounding by copollutants using multipollutant regression models.
These analyses indicate that although copollutant adjustment has varying degrees of influence
7-6
-------
Environmental and Health Impacts
on the SC>2 effect estimates, the effect of 862 on respiratory health outcomes appears to be
generally robust and independent of the effects of gaseous and particulate copollutants,
suggesting that the observed effects of 862 on respiratory endpoints occur independent of the
effects of other ambient air pollutants.
Consistent associations between short-term exposure to 862 and mortality have been
observed in epidemiologic studies, with larger effect estimates reported for respiratory
mortality than for cardiovascular mortality. While this finding is consistent with the
demonstrated effects of 862 on respiratory morbidity, uncertainty remains with respect to the
interpretation of these associations due to potential confounding by various copollutants. The
U.S. EPA has therefore concluded that the overall evidence is suggestive of a causal
relationship between short-term exposure to 862 and mortality. Significant associations
between short-term exposure to 862 and emergency department visits and hospital admissions
for cardiovascular diseases have also been reported. However, these findings have been
inconsistent across studies and do not provide adequate evidence to infer a causal relationship
between 862 exposure and cardiovascular morbidity.
7.1.1.3.3 Health Effects ofNOj
Information on the health effects of NC>2 can be found in the EPA Integrated Science
Assessment (ISA) for Nitrogen Oxides.41 The EPA has concluded that the findings of
epidemiologic, controlled human exposure, and animal toxicological studies provide evidence
that is sufficient to infer a likely causal relationship between respiratory effects and short-term
NC>2 exposure. The ISA concludes that the strongest evidence for such a relationship comes
from epidemiologic studies of respiratory effects including symptoms, emergency department
visits, and hospital admissions. The ISA also draws two broad conclusions regarding airway
responsiveness following NO2 exposure. First, the ISA concludes that NO2 exposure may
enhance the sensitivity to allergen-induced decrements in lung function and increase the
allergen-induced airway inflammatory response following 30-minute exposures of asthmatics
to NC>2 concentrations as low as 0.26 ppm. In addition, small but significant increases in non-
specific airway hyperresponsiveness were reported following 1-hour exposures of asthmatics
to 0.1 ppm NC>2. Second, exposure to NC>2 has been found to enhance the inherent
responsiveness of the airway to subsequent nonspecific challenges in controlled human
exposure studies of asthmatic subjects. Enhanced airway responsiveness could have
important clinical implications for asthmatics since transient increases in airway
responsiveness following NC>2 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 NC>2 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, NC>2 has also been linked to other
7-7
-------
Regulatory Impact Analysis
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
Information on the health effects of carbon monoxide (CO) can be found in the EPA
Integrated Science Assessment (ISA) for Carbon Monoxide.42 The ISA concludes that
ambient concentrations of CO are associated with a number of adverse health effects.E This
section provides a summary of the health effects associated with exposure to ambient
concentrations of CO.F
Human clinical studies of subjects with coronary artery disease show a decrease in the
time to onset of exercise-induced angina (chest pain) and electrocardiogram changes
following CO exposure. In addition, epidemiologic studies show associations between short-
term CO exposure and cardiovascular morbidity, particularly increased emergency room visits
and hospital admissions for coronary heart disease (including ischemic heart disease,
myocardial infarction, and angina). Some epidemiologic evidence is also available for
increased hospital admissions and emergency room visits for congestive heart failure and
cardiovascular disease as a whole. The ISA concludes that a causal relationship is likely
between short-term exposures to CO and cardiovascular morbidity. It also concludes that
available data are inadequate to conclude that a causal relationship exists between long-term
exposures to CO and cardiovascular morbidity.
Animal studies show various neurological effects with in-utero CO exposure.
Controlled human exposure studies report inconsistent neural and behavioral effects following
low-level CO exposures. The ISA concludes the evidence is suggestive of a causal
relationship with both short- and long-term exposure to CO and central nervous system
effects.
A number of epidemiologic and animal toxicological studies cited in the ISA have
evaluated associations between preterm birth and cardiac birth defects and CO exposure. The
epidemiologic studies provide limited evidence of a CO-induced effect on pre-term births and
E The ISA evaluates the health evidence associated with different health effects, assigning one of five "weight of
evidence" determination: causal relationship, likely to be a causal relationship, suggestive of a causal
relationship, inadequate to infer a causal relationship, and not likely to be a causal relationship. For definitions
of these levels of evidence, please refer to Section 1.6 of the ISA.
F Personal exposure includes contributions from many sources, and in many different environments. Total
personal exposure to CO includes both ambient and nonambient components; and both components may
contribute to adverse health effects.
7-8
-------
Environmental and Health Impacts
birth defects, with weak evidence for a decrease in birth weight. Animal toxicological studies
have found associations between perinatal CO exposure and decrements in birth weight, as
well as other developmental outcomes. The ISA concludes these studies are suggestive of a
causal relationship between long-term exposures to CO and developmental effects and birth
outcomes.
Epidemiologic studies provide evidence of effects on respiratory morbidity such as
changes in pulmonary function, respiratory symptoms, and hospital admissions associated
with ambient CO concentrations. A limited number of epidemiologic studies considered
copollutants such as ozone, SO2, and PM in two-pollutant models and found that CO risk
estimates were generally robust, although this limited evidence makes it difficult to
disentangle effects attributed to CO itself from those of the larger complex air pollution
mixture. Controlled human exposure studies have not extensively evaluated the effect of CO
on respiratory morbidity. Animal studies at levels of 50-100 ppm CO show preliminary
evidence of altered pulmonary vascular remodeling and oxidative injury. The ISA concludes
that the evidence is suggestive of a causal relationship between short-term CO exposure and
respiratory morbidity, and inadequate to conclude that a causal relationship exists between
long-term exposure and respiratory morbidity.
Finally, the ISA concludes that the epidemiologic evidence is suggestive of a causal
relationship between short-term exposures to CO and mortality. Epidemiologic studies
provide evidence of an association between short-term exposure to CO and mortality, but
limited evidence is available to evaluate cause-specific mortality outcomes associated with
CO exposure. In addition, the attenuation of CO risk estimates which was often observed in
copollutant models contributes to the uncertainty as to whether CO is acting alone or as an
indicator for other combustion-related pollutants. The ISA also concludes that there is not
likely to be a causal relationship between relevant long-term exposures to CO and mortality.
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.43 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.
-------
Regulatory Impact Analysis
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).
*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.
44
Noncancer health effects can result from chronic,0 subchronic,H or acute1 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
NATA, 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
0 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).
H Defined in the IRIS database as exposure to a substance spanning approximately 10% of the lifetime of an
organism.
1 Defined in the IRIS database as exposure by the oral, dermal, or inhalation route for 24 hours or less.
7-10
-------
Environmental and Health Impacts
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.45'46
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.47 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.48'49'50 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
51 52
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.53'54 The most sensitive noncancer effect observed in humans, based on current data,
is the depression of the absolute lymphocyte count in blood.55'56 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.57'58'59'60 EPA's IRIS program has not yet evaluated these new data.
7.1.1.5.2 1,3-Butadiene
EPA has characterized 1,3-butadiene as carcinogenic to humans by inhalation.61'62
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.63'64 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;
7-11
-------
Regulatory Impact Analysis
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.65
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.66 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. '68 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.69 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.70 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.71
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.72'73'74 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.7 '76'77'78 These findings are not supportive of interpreting the CUT model
results as providing a conservative (health protective) estimate of human risk. 9 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
Rfl
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
than previous IARC 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."81 EPA is reviewing the
7-12
-------
Environmental and Health Impacts
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.82'83
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.84 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.85'86 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.87 In short-term (4 week) rat studies, degeneration of
olfactory epithelium was observed at various concentration levels of acetaldehyde
OO OQ
exposure. ' 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.90 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.91 The IARC determined in 1995 that acrolein was not
classifiable as to its carcinogenicity in humans.92
Acrolein is extremely acrid and irritating to humans when inhaled, with acute
exposure resulting in upper respiratory tract irritation, mucus hypersecretion and congestion.
The intense irritancy of this carbonyl has been demonstrated during controlled tests in human
subjects, who suffer intolerable eye and nasal mucosal sensory reactions within minutes of
exposure.93 These data and additional studies regarding acute effects of human exposure to
7-13
-------
Regulatory Impact Analysis
acrolein are summarized in EPA's 2003 IRIS Human Health Assessment for acrolein.94
Evidence available from studies in humans indicate that levels as low as 0.09 ppm (0.21
mg/m3) for five minutes may elicit subjective complaints of eye irritation with increasing
concentrations leading to more extensive eye, nose and respiratory symptoms.95 Lesions to
the lungs and upper respiratory tract of rats, rabbits, and hamsters have been observed after
subchronic exposure to acrolein.96 Acute exposure effects in animal studies report bronchial
hyper-responsiveness.97 In a recent study, the acute respiratory irritant effects of exposure to
1.1 ppm acrolein were more pronounced in mice with allergic airway disease by comparison
to non-diseased mice which also showed decreases in respiratory rate.98 Based on these
animal data and demonstration of similar effects in humans (i.e., reduction in respiratory rate),
individuals with compromised respiratory function (e.g., emphysema, asthma) are expected to
be at increased risk of developing adverse responses to strong respiratory irritants such as
acrolein.
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.99'100 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.101
The draft reassessment completed external peer review.102 Based on external peer review
comments received, additional analyses are being undertaken. This external review draft does
not represent official agency opinion and was released solely for the purposes of external peer
review and public comment. The National Toxicology Program listed naphthalene as
"reasonably anticipated to be a human carcinogen" in 2004 on the basis of bioassays reporting
clear evidence of carcinogenicity in rats and some evidence of carcinogenicity in mice.103
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.104
7-14
-------
Environmental and Health Impacts
Naphthalene also causes a number of chronic non-cancer effects in animals, including
abnormal cell changes and growth in respiratory and nasal tissues.105
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 this final rule. Mobile
source air toxic compounds that would potentially be impacted include ethylbenzene,
propionaldehyde, toluene, and xylene. Information regarding the health effects of these
compounds can be found in EPA's IRIS database/
7.1.1.6 Exposure and Health Effects Associated with Traffic
Populations who live, work, or attend school near major roads experience elevated
exposure concentrations to a wide range of air pollutants, as well as higher risks for a number
of adverse health effects. While the previous sections of this RIA have focused on the health
effects associated with individual criteria pollutants or air toxics, this section discusses the
mixture of different exposures near major roadways, rather than the effects of any single
pollutant. As such, this section emphasizes traffic-related air pollution, in general, as the
relevant indicator of exposure rather than any particular pollutant.
Concentrations of many traffic-generated air pollutants are elevated for up to 300-500
meters downwind of roads with high traffic volumes.106 Numerous sources on roads
contribute to elevated roadside concentrations, including exhaust and evaporative emissions,
and resuspension of road dust and tire and brake wear. Concentrations of several criteria and
hazardous air pollutants are elevated near major roads. Furthermore, different semi-volatile
organic compounds and chemical components of particulate matter, including elemental
carbon, organic material, and trace metals, have been reported at higher concentrations near
major roads.
Populations near major roads experience greater risk of certain adverse health effects.
The Health Effects Institute published a report on the health effects of traffic-related air
pollution.107 It concluded that evidence is "sufficient to infer the presence of a causal
association" between traffic exposure and exacerbation of childhood asthma symptoms. The
HEI report also concludes that the evidence is either "sufficient" or "suggestive but not
sufficient" for a causal association between traffic exposure and new childhood asthma cases.
A review of asthma studies by Salam et al. (2008) reaches similar conclusions.108 The HEI
report also concludes that there is "suggestive" evidence for pulmonary function deficits
associated with traffic exposure, but concluded that there is "inadequate and insufficient"
J U.S. EPA Integrated Risk Information System (IRIS) database is available at: www.epa.gov/iris
7-15
-------
Regulatory Impact Analysis
evidence for causal associations with respiratory health care utilization, adult-onset asthma,
COPD symptoms, and allergy. A review by Holguin (2008) notes that the effects of traffic on
asthma may be modified by nutrition status, medication use, and genetic factors.109
The HEI report also concludes that evidence is "suggestive" of a causal association
between traffic exposure and all-cause and cardiovascular mortality. There is also evidence
of an association between traffic-related air pollutants and cardiovascular effects such as
changes in heart rhythm, heart attack, and cardiovascular disease. The HEI report
characterizes this evidence as "suggestive" of a causal association, and an independent
epidemiological literature review by Adar and Kaufman (2007) concludes that there is
"consistent evidence" linking traffic-related pollution and adverse cardiovascular health
outcomes.110
Some studies have reported associations between traffic exposure and other health
effects, such as birth outcomes (e.g., low birth weight) and childhood cancer. The HEI report
concludes that there is currently "inadequate and insufficient" evidence for a causal
association between these effects and traffic exposure. A review by Raaschou-Nielsen and
Reynolds (2006) concluded that evidence of an association between childhood cancer and
traffic-related air pollutants is weak, but noted the inability to draw firm conclusions based on
limited evidence.111
There is a large population in the U.S. living in close proximity of major roads.
According to the Census Bureau's American Housing Survey for 2007, approximately 20
million residences in the U.S., 15.6% of all homes, are located within 300 feet (91 m) of a
highway with 4+ lanes, a railroad, or an airport.112 Therefore, at current population of
approximately 309 million, assuming that population and housing similarly distributed, there
are over 48 million people in the U.S. living near such sources. The HEI report also notes that
in two North American cities, Los Angeles and Toronto, over 40% of each city's population
live within 500 meters of a highway or 100 meters of a major road. It also notes that about
33% of each city's population resides within 50 meters of major roads. Together, the
evidence suggests that a large U.S. population lives in areas with elevated traffic-related air
pollution.
People living near roads are often socioeconomically disadvantaged. According to the
2007 American Housing Survey, a renter-occupied property is over twice as likely as an
owner-occupied property to be located near a highway with 4+ lanes, railroad or airport. In
the same survey, the median household income of rental housing occupants was less than half
that of owner-occupants ($28,921/$59,886). Numerous studies in individual urban areas
report higher levels of traffic-related air pollutants in areas with high minority or poor
populations.113'114'115
Students may also be exposed in situations where schools are located near major
roads. In a study of nine metropolitan areas across the U.S., Appatova et al. (2008) found that
7-16
-------
Environmental and Health Impacts
on average greater than 33% of schools were located within 400 m of an Interstate, US, or
state highway, while 12% were located within 100 m.116 The study also found that among the
metropolitan areas studied, schools in the Eastern U.S. were more often sited near major
roadways than schools in the Western U.S.
Demographic studies of students in schools near major roadways suggest that this
population is more likely than the general student population to be of non-white race or
Hispanic ethnicity, and more often live in low socioeconomic status locations. 117>118'119 There
is some inconsistency in the evidence, which may be due to different local development
patterns and measures of traffic and geographic scale used in the studies.116
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: particulate 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
emissions of primary PM2.5 and secondary PM2.5 precursors such as NOx- 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.
EPA is pursuing a two-part strategy to address visibility. First, EPA has concluded
that PM2.5 causes adverse effects on visibility in various locations, depending on PM
concentrations and factors such as chemical composition and average relative humidity, and
has set secondary PM2.5 standards.K The secondary PM2.5 standards act in conjunction with
the regional haze program. The regional haze rule (64 FR 35714) was put in place in July
1999 to protect the visibility in mandatory class I federal areas. There are 156 national parks,
forests and wilderness areas categorized as mandatory class I federal areas (62 FR 38680-81,
July 18, 1997).L Visibility can be said to be impaired in both PM2.5 nonattainment areas and
mandatory class I federal areas. Figure 7-1 shows the location of the 156 mandatory class I
federal areas.
K The existing annual primary and secondary PM2.5 standards have been remanded and are being addressed in
the currently ongoing PM NAAQS review.
L These areas are defined in CAA section 162 as those national parks exceeding 6,000 acres, wilderness areas
and memorial parks exceeding 5,000 acres, and all international parks which were in existence on August 7,
1977.
7-17
-------
Regulatory Impact Analysis
Produced by NFS Air Resources
* Rainbow Lake, Wl and Br adwell Bay, FL are Class 1 Areas
where visibility is not an important air quality related value
Figure 7-1 Mandatory Class I Areas in the U.S.
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 1 10 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 PMi0 and
PMi.smass, and for key constituents of PM^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
7-18
-------
Environmental and Health Impacts
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. In addition to
this indirect method of assessing light extinction, there are optical measurements which
directly measure light extinction or its components. Such measurements are taken principally
with either a transmissometer, which measures total light extinction, or by combining the PM
light scattering measured by integrating nephelometers with the PM light absorption measured
by an aethalometer. Scene characteristics are typically recorded three times daily with 35
millimeter photography and are used to determine the quality of visibility conditions (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. Visibility is
typically worse in the summer months and the rural East generally has higher levels of
impairment than remote sites in the West. Figures 9-9 through 9-11 in the PM ISA detail the
percent contributions to particulate light extinction for ammonium nitrate and sulfate, EC and
OC, and coarse mass and fine soil, by season.120
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.121 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".122 Like carbon dioxide (CO2)
and other gaseous substances, ozone enters plant tissues primarily through apertures (stomata)
in leaves in a process called "uptake".123 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.124'125 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
reduced,126 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
7-19
-------
Regulatory Impact Analysis
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.127'128
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)129'130'131 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.132
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.133'134 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.135'136
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
7-20
-------
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.137 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.138'139'140 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."141 In addition, economic studies have shown
reduced economic benefits as a result of predicted reductions in crop yields associated with
observed ozone levels.142'143'144
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.145 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.146 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.147
In the U.S. this indicator is based on data from the U.S. Department of Agriculture
(USDA) 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.148'149 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
7-21
-------
Regulatory Impact Analysis
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.
7.1.2.2.1 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.150'151 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 indicator 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.152'153
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 snowed 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.
7-22
-------
Environmental and Health Impacts
Region 1
(54 sites)
Region 2
[42 sites)
Region 3
(111 sites)
Region 4
(227 sites)
Region 5
Region 6
(59 sites'i
Region 7
(A3 sites)
Region 8
(72 sites'i
Region 9
(80 sites)
Region 10
(57 sites)
^Coverage:
located in A
:7>3tals may
rounding.
" j ' j ssturci
2006
legree ol injury:
None Low Moderate
High Savsre
DBreent nl monitoring Bites in each category:
68.5
61.0
16.7 -1.1
-3.7
21.4 71 7
E5.9 mo 14.4 7
753
10.1 7 ] "
.12.4
.24.5
-3.E
HI
1BJ 5
94.9
85.7
-5.1
9.5
- 3.2
•1.6
100.0
76.3
12.5 fl.8
Ei
100.0
J45 monitoring sites, EPA Regions
1 StHtSS.
nat add tu 100% due ta
%
R- ' , " T •' r-,r^.-J Qfj-iAvv.
^
|® V—
c^
O
\
©»l
Figure 7-2 Ozone Injury to Forest Plants in U.S. by EPA Regions, 2002
ab
7-23
-------
Regulatory Impact Analysis
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
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.154 Ozone may have other adverse impacts on plants (e.g., reduced
productivity) that do not show signs of visible foliar injury.155
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.156
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-
7-24
-------
Environmental and Health Impacts
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.15?
Some of the common tree species in the United States that are sensitive to ozone are
black cherry (Primus serotind), 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 (Pinus
ponderosa). 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
Particulate 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). The following characterizations of the nature of these environmental
effects are based on information contained in the 2009 PM ISA and the 2005 PM Staff
Paper.158'159
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 acidification, nutrient enrichment, and eutrophication of aquatic
ecosystems.160
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
7-25
-------
Regulatory Impact Analysis
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 particulate 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
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.161 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.162 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 condition known as nitrogen saturation. Nitrogen
saturation is the condition in which nitrogen inputs from atmospheric deposition and other
sources exceed the biological requirements of the ecosystem. The effects associated with
nitrogen saturation 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.163
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
7-26
-------
Environmental and Health Impacts
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
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 NOAA 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.164
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 impacting forest growth.165 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. 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
7-27
-------
Regulatory Impact Analysis
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, exert toxic effects on the plant itself, 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. 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.166 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. As the fallen leaves
decompose, the heavy metals are transferred into the soil.167'168 Upon entering the soil
environment, PM pollutants can alter ecological processes of energy flow and nutrient
cycling, inhibit nutrient uptake, change ecosystem structure, and affect ecosystem
biodiversity. Many of the most important effects occur in the soil. The soil environment is one
of the most dynamic sites of biological interaction in nature. It is inhabited by microbial
communities of bacteria, fungi, and actinomycetes. These organisms are essential participants
in the nutrient cycles that make elements available for plant uptake. Changes in the soil
environment that influence the role of the bacteria and fungi in nutrient cycling determine
plant and ultimately ecosystem response.169
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.170'171 Over
fifty percent of the mercury in the Chesapeake Bay has been attributed to atmospheric
deposition.172 Overall, the National Science and Technology Council identifies atmospheric
deposition as the primary source of mercury to aquatic systems.173 Forty-four states have
issued health advisories for the consumption offish 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
7-28
-------
Environmental and Health Impacts
use.174'175 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.176 Plant uptake of platinum has been
observed at these locations.
7.1.2.4.3 Deposition ofPolycyclic Organic Matter
Polycyclic 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.177 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 particulate matter. Since the majority of PAHs are
adsorbed onto particles less than 1.0 j^m in diameter, long range transport is possible.
However, studies have shown that PAH compounds adsorbed onto diesel exhaust particulate
and exposed to ozone have half lives of 0.5 to 1.0 hours.178
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.17 >18° Analyses of PAH deposition in
Chesapeake and Galveston Bay indicate that dry deposition and gas exchange from the
1 R1 1 R9 «-'«—'
atmosphere to the surface water predominate. ' 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.183 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.184 Thus, dry deposition is likely the main
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.185 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.186
Cousins et al. estimate that more than ninety percent of semi-volatile organic
compound (SVOC) emissions in the United Kingdom deposit on soil.187 An analysis of PAH
concentrations near a Czechoslovakian roadway indicated that concentrations were thirty
times greater than background.188
7-29
-------
Regulatory Impact Analysis
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.189 In
laboratory experiments, a wide range of tolerance to VOCs has been observed.190 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.191
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.192'193'194
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.
7.2 Non-GHG Air Quality Impacts
This section presents the methodology and results of EPA's air quality modeling to
determine the projected impact of the vehicle standards finalized in this rule on ambient
concentrations of criteria and air toxic pollutants. Section 7.1 above describes the health and
environmental effects associated with the criteria and air toxic pollutants that are impacted by
this rule, and Section 7.3 describes the methodology for calculating monetized benefits due to
reductions in adverse health effects associated with PM2.5 and ozone.
7-30
-------
Environmental and Health Impacts
7.2.1 Air Quality Modeling Methodology
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 - local, regional, national, and global. This
section provides detailed information on the photochemical model used for our air quality
analysis (the Community Multi-scale Air Quality (CMAQ) model), atmospheric reactions and
the role of chemical mechanisms in modeling, and model uncertainties and limitations.
Further discussion of the modeling methodology is included in the Air Quality Modeling
Technical Support Document (AQM TSD) found in the docket for this rule. Results of the air
quality modeling are presented in Section 7.2.2.
7.2.1.1 Modeling Methodology
A national-scale air quality modeling analysis was performed to estimate future year
annual PM2.5 concentrations, 24-hour PM2.5 concentrations, 8-hour ozone concentrations, air
toxics concentrations, and nitrogen and sulfur deposition levels for future years. The 2005-
based CMAQ modeling platform was used as the basis for the air quality modeling of the
future reference case and the future control scenario for this final rule. This platform
represents a structured system of connected modeling-related tools and data that provide a
consistent and transparent basis for assessing the air quality response to projected changes in
emissions. The base year of data used to construct this platform includes emissions and
meteorology for 2005. The platform was developed by the U.S. EPA's Office of Air Quality
Planning and Standards in collaboration with the Office of Research and Development and is
intended to support a variety of regulatory and research model applications and analyses.
The CMAQ modeling system is a non-proprietary, publicly available, peer-reviewed,
state-of-the-science, three-dimensional, grid-based Eulerian air quality grid model designed to
estimate the formation and fate of oxidant precursors, primary and secondary PM
concentrations, acid deposition, and air toxics, over regional and urban spatial scales for given
input sets of meteorological conditions and emissions.195'196'197 The CMAQ model version 4.7
was most recently peer-reviewed in February of 2009 for the U.S. EPA.M The CMAQ model
M Report on the peer-review is still being finalized. Draft available upon request from Director S.T.Rao,
Atmospheric Modeling and Analysis Division; rao.st@epa.gov; 919-541-4541. Allen, D., Burns, D., Chock, D.,
Kumar, N., Lamb, B., Moran, M. (February 2009 Draft Version). Report on the Peer Review of the Atmospheric
Modeling and Analysis Division, NERL/ORD/EPA. U.S. EPA, Research Triangle Park, NC.
7-31
-------
Regulatory Impact Analysis
is a well-known and well-respected tool and has been used in numerous national and
international applications.198'199'200 This 2005 multi-pollutant modeling platform used CMAQ
version 4.7.1N with a minor internal change made by the U.S. EPA CMAQ model developers
intended to speed model runtimes when only a small subset of toxics species are of interest.
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 used CMAQ v4.7.1 which reflects updates to version 4.7 to improve the
underlying science. These include aqueous chemistry mass conservation improvements,
improved vertical convective mixing and lowered CB05 mechanism unit yields for acrolein
from 1,3-butadiene tracer reactions which were updated to be consistent with laboratory
measurements. Section 7.2.1.2.2 of this RIA discusses the chemical mechanism and SOA
formation.
7.2.1.1.1 Model Domain and Configuration
The CMAQ modeling domain encompasses all of the lower 48 States and portions of
Canada and Mexico. The modeling domain is made up of a large continental U.S. 36
kilometer (km) grid and two 12 km grids (an Eastern US and a Western US domain), as
shown in Figure 7-3. The modeling domain contains 14 vertical layers with the top of the
modeling domain at about 16,200 meters, or 100 millibars (mb).
N CMAQ version 4.7 was released on December, 2008. It is available from the Community Modeling and
Analysis System (CMAS) as well as previous peer-review reports at: http://www.cmascenter.org.
7-32
-------
Environmental and Health Impacts
12km Western Domain (WRAP)
origin .2112000. .272000
'
12km Eastern Domain
Figure 7-3 Map of the CMAQ Modeling Domain
7.2.1.1.2 Model Inputs
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 were derived from simulations of the Pennsylvania State
University/National Center for Atmospheric Research Mesoscale Model201 for the entire year
of 2005 over model domains that are slightly larger than those shown in Figure 7-3. 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.202 The meteorology for the national 36 km grid and the two 12 km
grids were developed by EPA and are described in more detail within the AQM TSD. The
meteorological outputs from MM5 were processed to create model-ready inputs for CMAQ
using the Meteorology-Chemistry Interface Processor (MCIP) version 3.4, for example:
horizontal wind components (i.e., speed and direction), temperature, moisture, vertical
diffusion rates, and rainfall rates for each grid cell in each vertical layer.
203
The lateral boundary and initial species concentrations are provided by a three-
dimensional global atmospheric chemistry model, the GEOS-CHEM model.204 The global
GEOS-CHEM model simulates atmospheric chemical and physical processes driven by
7-33
-------
Regulatory Impact Analysis
assimilated meteorological observations from the NASA's Goddard Earth Observing System
(GEOS). This model was run for 2005 with a grid resolution of 2 degree x 2.5 degree
(latitude-longitude) and 20 vertical layers. The predictions were 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
were used as the initial/boundary state for all subsequent 12 km finer grid modeling.
The emissions inputs used for the 2005 base year and each of the future year base
cases and control scenarios analyzed for this rule are summarized in Chapter 5 of this RIA.
7.2.1.1.3 CMAQ Evaluation
An operational model performance evaluation for ozone, PM2.5 and its related
speciated components (e.g., sulfate, nitrate, elemental carbon, organic carbon, etc.), nitrate
and sulfate deposition, and specific air toxics (formaldehyde, acetaldehyde, benzene, 1,3-
butadiene, and acrolein) was conducted using 2005 state/local monitoring data in order to
estimate the ability of the CMAQ modeling system to replicate base year concentrations.
Model performance statistics were calculated for observed/predicted pairs of
daily/monthly/seasonal/annual concentrations. Statistics were generated for the following
geographic groupings: domain wide, Eastern vs. Western (divided along the 100th meridian),
and each Regional Planning Organization (RPO) region.0 The "acceptability" of model
performance was judged by comparing our results to those found in recent regional PM2.5
model applications for other, non-EPA studies.p Overall, the performance for the 2005
modeling platform is within the range or close to that of these other applications. The
performance of the CMAQ modeling was evaluated over a 2005 base case. The model was
able to reproduce historical concentrations of ozone and PM2.5 over land with low bias and
error results. Model predictions of annual formaldehyde, acetaldehyde and benzene showed
relatively small bias and error results when compared to observations. The model yielded
larger bias and error results for 1,3 butadiene and acrolein based on limited monitoring sites.
A more detailed summary of the 2005 CMAQ model performance evaluation is available
within the AQM TSD found in the docket of this rule.
0 Regional Planning Organization regions include: Mid-Atlantic/Northeast Visibility Union (MANE-VU),
Midwest Regional Planning Organization - Lake Michigan Air Directors Consortium (MWRPO-LADCO),
Visibility Improvement State and Tribal Association of the Southeast (VISTAS), Central States Regional Air
Partnership (CENRAP), and Western Regional Air Partnership (WRAP).
p These other modeling studies represent a wide range of modeling analyses which cover various models, model
configurations, domains, years and/or episodes, chemical mechanisms, and aerosol modules.
7-34
-------
Environmental and Health Impacts
7.2.1.1.4 Model Simulation Scenarios
As part of our analysis for this rulemaking, the CMAQ modeling system was used to
calculate daily and annual PM2.5 concentrations, 8-hour ozone concentrations, annual and
seasonal (summer and winter) air toxics concentrations, and annual nitrogen and sulfur
deposition total levels for each of the following emissions scenarios:
- 2005 base year
- 2030 reference case projection
- 2030 control case projection
The emission inventories used in the air quality and benefits modeling are different
from the final rule inventories due to the considerable length of time required to conduct the
modeling. However, the air quality modeling inventories are generally consistent with the
final emission inventories, so the air quality modeling adequately reflects the effects of the
rule. The emission inventories used for air quality modeling are discussed in Section 5.8 of
this RIA. The emissions modeling TSD, found in the docket for this rule (EPA-HQ-OAR-
2009-0472), contains a detailed discussion of the emissions inputs used in our air quality
modeling.
We use the predictions from the model in a relative sense by combining the 2005 base-
year predictions with predictions from each future-year scenario and applying these modeled
ratios to ambient air quality observations to estimate daily and annual PM2.5 concentrations,
and 8-hour ozone concentrations for each of the 2030 scenarios. The ambient air quality
observations are average conditions, on a site-by-site basis, for a period centered around the
model base year (i.e., 2003-2007).
The projected daily and annual PM2.5 design values were calculated using the
Speciated Modeled Attainment Test (SMAT) approach. The SMAT uses a Federal Reference
Method (FRM) mass construction methodology that results in reduced nitrates (relative to the
amount measured by routine speciation networks), higher mass associated with sulfates
(reflecting water included in FRM measurements), and a measure of organic carbonaceous
mass that is derived from the difference between measured PM2.5 and its non-carbon
components. This characterization of PM2.5 mass also reflects crustal material and other
minor constituents. The resulting characterization provides a complete mass balance. It does
not have any unknown mass that is sometimes presented as the difference between measured
PM2.5 mass and the characterized chemical components derived from routine speciation
measurements. However, the assumption that all mass difference is organic carbon has not
been validated in many areas of the U.S. The SMAT methodology uses the following PM2.5
species components: sulfates, nitrates, ammonium, organic carbon mass, elemental carbon,
crustal, water, and blank mass (a fixed value of 0.5 (^g/m3). More complete details of the
7-35
-------
Regulatory Impact Analysis
SMAT procedures can be found in the report "Procedures for Estimating Future PM2.5 Values
for the CAIR Final Rule by Application of the (Revised) Speciated Modeled Attainment Test
(SMAT)".205 For this latest analysis, several datasets and techniques were updated. These
changes are fully described within the technical support document for the Small SI Engine
Rule modeling AQM TSD.206 The projected 8-hour ozone design values were calculated
using the approach identified in EPA's guidance on air quality modeling attainment
demonstrations.207
Additionally, we conducted an analysis to compare the absolute and percent
differences between the 2030 control case and the 2030 reference cases for annual and
seasonal ethanol and five air toxics of interest (formaldehyde, acetaldehyde, benzene, 1,3-
butadiene, and acrolein), as well as annual nitrate and sulfate deposition. These data were not
compared in a relative sense due to the limited observational data available.
7.2.1.2 Chemical Mechanisms in Modeling
This rule presents inventories for NOx, VOC, CO, PMi.5, 862, NHs, and five air
toxics: benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and acrolein. The five air toxics
are explicit model species in the CMAQv4.7 model with carbon bond 5 (CB05)
("J(-\Q
mechanisms. In addition to direct emissions, photochemical processes mechanisms are
responsible for formation of some of these compounds in the atmosphere from precursor
emissions. For some pollutants such as PM, formaldehyde, and acetaldehyde, many
photochemical processes are involved. CMAQ therefore also requires inventories for a large
number of other air toxics and precursor pollutants. Methods used to develop the air quality
inventories can be found in Chapter 5 of the RIA.
In the CB05 mechanism, the chemistry of thousands of different VOCs in the
atmosphere are represented by a much smaller number of model species which characterize
the general behavior of a subset of chemical bond types; this condensation is necessary to
allow the use of complex photochemistry in a fully 3-D air quality model.209
Complete combustion of ethanol in fuel produces carbon dioxide (CO2) and water
(F^O). Incomplete combustion results in the production of other air pollutants, such as
acetaldehyde and other aldehydes, and the release of unburned ethanol. Ethanol is also
present in evaporative emissions. In the atmosphere, ethanol from unburned fuel and
evaporative emissions can undergo photodegradation to form aldehydes (acetaldehyde and
formaldehyde) and peroxyacetyl nitrate (PAN), and also plays a role in ground-level ozone
formation. Mechanisms for these reactions are included in CMAQ. Additionally, other
aromatic hydrocarbons (ARC) and hydrocarbons are considered because any increase in
acetyl peroxy radicals due to ethanol increases might be counterbalanced by a decrease in
radicals resulting from decreases in ARC and other hydrocarbons.
7-36
-------
Environmental and Health Impacts
CMAQ includes 63 inorganic reactions to account for the cycling of all relevant
oxidized nitrogen species and cycling of radicals, including the termination of NO2 and
formation of nitric acid (HNO3) without PAN formation.12
NO2 + -OH + M -> HNO3 + M k = 1.19 x 10"11 cn^molecule'Y1 21°
The CB05 mechanism also includes more than 90 organic reactions that include
alternate pathways for the formation of acetyl peroxy radical, such as by reaction of
methylglyoxal, which is also formed from reactions of ARC. Alternate reactions of acetyl
peroxy radical, such as oxidation of NO to form NO2, which again leads to ozone formation,
are also included.
Atmospheric reactions and chemical mechanisms involving several key formation
pathways are discussed in more detail in the following sections.
7.2.1.2.1 Acetaldehyde
Acetaldehyde is the main photodegradation product of ethanol, as well as other
precursor hydrocarbons. Acetaldehyde is also a product of fuel combustion. In the
atmosphere, acetaldehyde can react with the OH radical and O2 to form the acetyl peroxy
radical [CH3C(O)OO-].R This radical species can then further react with nitric oxide (NO), to
produce formaldehyde (HCHO), or with nitrogen dioxide (NO2), to produce PAN
[CH3C(O)OONO2]. An overview of these reactions and the corresponding reaction rates are
provided below.s
CH3CHO + -OH -> CH3C-O + H2O k = 1.5 x 10"11 cm3molecule V1 211
CH3C-O + O2 + M -> CH3C(O)OO- + M
CH3C(O)OO- + NO -> CH3C(O)O- + NO2 k = 2.0 x 10"11 cnrWlecule'V1 212
CH3C(O)O- -> -CH3 + CO2
•CH3 + O2 + M -> CH3OO- + M
CH3OO- + NO -> CH3O- + NO2
Q All rate coefficients in this RIA are listed at 298 K and, if applicable, 1 bar of air.
R Acetaldehyde is not the only source of acetyl peroxy radicals in the atmosphere. For example, dicarbonyl
compounds (methylglyoxal, biacetyl, and others) also form acetyl radicals, which can further react to form
peroxyacetyl nitrate (PAN).
s All rate coefficients in this RIA are listed at 298 K and, if applicable, 1 bar of air.
7-37
-------
Regulatory Impact Analysis
CH3O + O2 -> HCHO + HO2
CH3C(O)OO- + NO2 + M -> CH3C(O)OONO2 + M k = 1.0 x 10"11 cn^molecule'Y1 213
Acetaldehyde can also photolyze (hv), which predominantly produces -CH3 and HCO:
CH3CHO + hv -> -CH3 + HCO X = 240-380 nm 214
As mentioned above, -CH3 is oxidized in the atmosphere to produce formaldehyde
(HCHO). Formaldehyde is also a product of hydrocarbon combustion. In the atmosphere,
formaldehyde undergoes photolysis and reaction with the OH radical, NO3 radical, and ozone,
and the resulting lifetimes are ~4 hours, 1.2 days, 83 days, and >4.5 years, respectively.1
Formaldehyde is removed mainly by photolysis whereas the higher aldehydes, those with two
or more carbons such as acetaldehyde, react predominantly with OH radicals. The photolysis
of formaldehyde is a source of additional radicals, and as shown above, these radicals can
react with NO2 to form PAN in the atmosphere.
HCHO + hv -> H + HCO X = 240-360 nm
215
CB05 mechanisms for acetaldehyde formation warrant a detailed discussion given the
increase in vehicle and engine exhaust emissions for this pollutant and ethanol, which can
form acetaldehyde in the air. Acetaldehyde is represented explicitly in the CB05 chemical
mechanism216'217 by the ALD2 model species, which can be both formed from other VOCs
and can decay via reactions with oxidants and radicals. The reaction rates for acetaldehyde,
as well as for the inorganic reactions that produce and cycle radicals, and the representative
reactions of other VOCs have all been updated to be consistent with recommendations in the
literature.218
The decay reactions of acetaldehyde are fewer in number and can be characterized
well because they are explicit representations. Acetaldehyde can photolyze in the presence of
sunlight or react with molecular oxygen (O3(P)), hydroxyl radical (OH), or nitrate radicals.
Of these reactions, both photolysis and reaction with OH are the most important reactions
determining loss of acetaldehyde. The reaction rates are based on expert recommendations,219
and the photolysis rate is from IUPAC recommendations.
In CMAQ v4.7, the acetaldehyde that is formed from photochemical reactions is
tracked separately from that which is due to direct emission and transport of direct emissions.
In CB05, there are 25 different reactions that form acetaldehyde in molar yields ranging from
T Lifetime calculated using the following: for photolysis, with overhead sun (at noontime during the summer);
for OH radical reactions, a 12-hour daytime average of 2.0 x 106 molecule cm"3; for NO3 radical reactions, a 12-
hour nighttime average of 5 x 108 molecule cm"3; and for ozone, a 24-hour average of 7 x 1011 molecule cm"3.
7-38
-------
Environmental and Health Impacts
0.02 (ozone reacting with lumped products from isoprene oxidation) to 2.0 (cross reaction of
acylperoxy radicals, CXO3). The specific parent VOCs that contribute the most to
acetaldehyde concentrations vary spatially and temporally depending on characteristics of the
ambient air, but alkenes in particular are found to play a large role. The IOLE model species,
which represents internal carbon-carbon double bonds, has high emissions and relatively high
yields of acetaldehyde. The OLE model species, representing terminal carbon double bonds,
also plays a role because it has high emissions although lower acetaldehyde yields.
Production from peroxyproprional nitrate and other peroxyacylnitrates (PANX) and aldehydes
with 3 or more carbon atoms also play an important role. Thus, the amount of acetaldehyde
(and formaldehyde as well) formed in the ambient air as well as emitted in the exhaust (the
latter being accounted for in emission inventories) is affected by changes in these precursor
compounds due to the addition of ethanol to fuels (e.g., decreases in alkenes would cause
some decrease of acetaldehyde, and to a larger extent, formaldehyde).
The reaction of ethanol (CHsC^OH) with OH is slower than some other important
reactions but can be an important source of acetaldehyde if the emissions are large. Based on
kinetic data for molecular reactions, the only important chemical loss process for ethanol (and
other alcohols) is reaction with the hydroxyl radical (-OH).220 This reaction produces
acetaldehyde (CHsCHO) with a 90% yield.221 The lifetime of ethanol in the atmosphere can
be calculated from the rate coefficient, k, and due to reaction with the OH radical, occurs on
the order of a day in polluted urban areas or several days in unpolluted areas.u
CH3CH2OH + -OH -> CH3C-HOH + H2O k = 3.2 x 10"12 cn^molecule'Y1 222
CH3C-HOH + O2 -> CH3CHO + HO2
In CB05, reaction of one molecule of ethanol yields 0.90 molecules of acetaldehyde.
It assumes the majority of the reaction occurs through H-atom abstraction of the more
weakly-bonded methylene group, which reacts with oxygen to form acetaldehyde and
hydroperoxy radical (HO2), and the remainder of the reaction occurs at the -CH3 and -OH
groups, creating formaldehyde (HCHO), oxidizing NO to NO2 (represented by model species
XO2) and creating glycoaldehyde, which is represented as ALDX:
CH3CHOH + OH -> HO2 + 0.90 CH3CHO + 0.05 ALDX + 0.10 HCHO + 0.10 XO2
7.2.7.2.2 Secondary Organic Aerosols
Secondary organic aerosol (SOA) chemistry research described below has led to
implementation of new pathways for secondary organic aerosol (SOA) in CMAQ 4.7, based
on recommendations of Edney et al. and the recent work of Carlton et al.223'224 In previous
u All rate coefficients in this RIA are listed at 298 K and, if applicable, 1 bar of air.
7-39
-------
Regulatory Impact Analysis
versions of the CMAQ model, all SOA was treated as semi-volatile, whereas in CMAQ v4.7,
non-volatile SOA are simulated as well, including SOA originating from aromatic oxidation
under low-NOx conditions.
7.2.1.2.2.1 SOA Research
SOA results when products of atmospheric transformation or photooxidation of a
volatile organic compound (VOC) form or partition to the particle phase. Current research
suggests SOA contributes significantly to ambient organic aerosol (OA) concentrations, and
in Southeast and Midwest States may make up more than 50% (although the contribution
varies from area to area) of the organic fraction of PMi.5 during the summer (but less in the
winter).225'226 A wide range of laboratory studies conducted over the past twenty years show
that anthropogenic aromatic hydrocarbons and long-chained alkanes, along with biogenic
isoprene, monoterpenes, and sesquiterpenes, contribute to SOA formation.227'228'229'2 °'231
Anthropogenic SOA is a small portion of all SOA; most is biogenic and varies with season.
Based on these laboratory results, SOA chemical mechanisms have been developed and
integrated into air quality models such as the CMAQ model and have been used to predict OA
232
concentrations.
Over the past 10 years, ambient OA concentrations have been routinely measured in
the U.S. and some of these data have been used to determine, by employing source/receptor
methods, the contributions of the major OA sources, including biomass burning and vehicular
gasoline and diesel exhaust. Since mobile sources are a significant source of VOC emissions,
OT?
currently accounting for approximately 50% of anthropogenic VOC, mobile sources are
also an important source of SOA.
Toluene is an important contributor to anthropogenic SOA. Other aromatic
compounds contribute as well, but the extent of their contribution has not yet been quantified.
Mobile sources are the most significant contributor to ambient toluene concentrations as
shown by analyses done for the 2002 National Air Toxics Assessment (NATA)234 and the
Mobile Source Air Toxics (MSAT) Rule.235 2002 NATA indicates that onroad and nonroad
mobile sources accounted for 70% (2.24 |^g/m3) of the total average nationwide ambient
concentration of toluene (3.24 |^g/m3), when the contribution of the estimated "background" is
apportioned among source sectors.
The amount of toluene in gasoline influences the amount of toluene emitted in vehicle
exhaust and evaporative emissions, although, like benzene, some toluene is formed in the
combustion process. In turn, levels of toluene and other aromatics in gasoline are potentially
influenced by the amount of ethanol blended into the fuel. Due to the high octane quality of
ethanol, it greatly reduces the need for and levels of other high-octane components such as
aromatics including toluene (which is the major aromatic compound in gasoline). Since
toluene contributes to SOA and the toluene level of gasoline is decreasing, it is important to
assess the effect of these reductions on ambient PM.
7-40
-------
Environmental and Health Impacts
It is unlikely that ethanol would directly form SOA or affect SOA formation indirectly
through changes in the radical populations from increasing ethanol exhausts. Nevertheless,
scientists at the U.S. EPA's Office of Research and Development's National Exposure
Research Laboratory recently directed experiments to investigate ethanol's SOA forming
potential.236 The experiments were conducted under conditions where peroxy radical
reactions would predominate (irradiations performed in the absence of NOx and OH produced
from the photolysis of hydrogen peroxide). This was the most likely scenario under which
SOA formation could occur, since a highly oxygenated C4 organic would be potentially
made. As expected, no SOA was produced. From these experiments, the upper limit for the
aerosol yield would have been less than 0.01% based on scanning mobility particle sizer
(SMPS) data. Given the expected negative result based on these initial smog chamber
experiments, these data were not published.
In general, a review of the literature shows limited data on SOA concentrations,
largely due to the lack of analytical methods for identifying and determining the
concentrations of the highly polar organic compounds that make up SOA. The most widely
applied method of estimating total ambient SOA concentrations is the EC tracer method using
OT7 9^8
ambient data which estimates of the OC/EC ratio in primary source emissions. ' SOA
concentrations have also been estimated using OM (organic mass) to OC (organic carbon)
ratios, which can indicate that SOA formation has occurred, or by subtracting the
source/receptor-based total primary organic aerosol (POA) from the measured OC
concentration.239 Such methods, however, may not be quantitatively accurate and provide no
information on the contribution of individual biogenic and anthropogenic SOA sources, which
is critical information needed to assess the impact of specific sources and the associated
health risk. These methods assume that OM containing additional mass from oxidation of OC
comes about largely (or solely) from SOA formation. In particular, the contributions of
anthropogenic SOA sources, including those of aromatic precursors, are required to determine
exposures and risks associated with replacing fossil fuels with biofuels.
Upon release into the atmosphere, numerous VOC compounds can react with free
radicals in the atmosphere to form SOA. While this has been investigated in the laboratory,
there is relatively little information available on the specific chemical composition of SOA
compounds themselves from specific VOC precursors. This absence of compositional data
from the precursors has largely prevented the identification of aromatically-derived SOA in
ambient samples which, in turn, has prevented observation-based measurements of the
aromatic and other SOA contributions to ambient PM levels.
As a first step in determining the ambient SOA concentrations, EPA has developed a
tracer-based method to estimate such concentrations.240'241 The method is based on using
mass fractions of SOA tracer compounds, measured in smog chamber-generated SOA
samples, to convert ambient concentrations of SOA tracer compounds to ambient SOA
concentrations. This method consists of irradiating the SOA precursor of interest in a smog
chamber in the presence of NOX, collecting the SOA produced on filters, and then analyzing
7-41
-------
Regulatory Impact Analysis
the samples for highly polar compounds using advanced analytical chemistry methods.
Employing this method, candidate tracers have been identified for several VOC compounds
which are emitted in significant quantities and known to produce SOA in the atmosphere.
Some of these SOA-forming compounds include toluene, a variety of monoterpenes, isoprene,
and p-caryophyllene, the latter three of which are emitted by vegetation and are more
significant sources of SOA than toluene. Smog chamber work can also be used to investigate
949 94.^ 943T94S
SOA chemical formation mechanisms. ''
Although these concentrations are only estimates, due to the assumption that the mass
fractions of the smog chamber SOA samples using these tracers are equal to those in the
ambient atmosphere, there are presently no other means available for estimating the SOA
concentrations originating from individual SOA precursors. Among the tracer compounds
observed in ambient PM2.5 samples are two tracer compounds that have been identified in
smog chamber aromatic SOA samples.246 To date, these aromatic tracer compounds have
been identified, in the laboratory, for toluene and m-xylene SOA. Additional work is
underway by the EPA to determine whether these tracers are also formed by benzene and
other alkylbenzenes (including o-xylene, p-xylene, 1,2,4-trimethylbenzene, and
ethylbenzene).
One caveat regarding this work is that a large number of VOCs emitted into the
atmosphere, which have the potential to form SOA, have not yet been studied in this way. It
is possible that these unstudied compounds produce SOA species which are being used as
tracers for other VOCs. This means that the present work could overestimate the amount of
SOA formed in the atmosphere by the VOCs studied to date. This approach may also
estimate entire hydrocarbon classes (e.g., all methylsubstituted-monoaromatics or all
monoterpenes) and not individual precursor hydrocarbons. Thus the tracers could be broadly
representative and not indicative of individual precursors. This is still unknown. Also,
anthropogenic precursors play a role in formation of atmospheric radicals and aerosol acidity,
and these factors influence SOA formation from biogenic hydrocarbons. This anthropogenic
and biogenic interaction, important to EPA and others, needs further study. The issue of SOA
formation from aromatic precursors is an important one to which EPA and others are paying
significant attention. For benzene, smog chamber studies show that benzene forms SOA
possibly through reactions with NOX. Early smog chamber work suggests benzene might be
relatively inert in forming SOA, although this study may not be conclusive.247 However,
more recent work shows that benzene does form SOA in smog chambers.248'249 This new
smog chamber work shows that benzene can be oxidized in the presence of NOx to form SOA
with maximum mass of SOA being 8-25% of the mass of benzene. As mentioned above,
work is needed to determine if a tracer compound can be found for benzene SOA which might
indicate how much of ambient SOA comes from benzene.
The aromatic tracer compounds and their mass fractions have also been used to
estimate monthly ambient aromatic SOA concentrations from March 2004 to February 2005
in five U.S. Midwestern cities.250 The annual tracer-based SOA concentration estimates were
7-42
-------
Environmental and Health Impacts
0.15, 0.18, 0.13, 0.15, and 0.19 jig carbon/m3 for Bondville, IL, East St. Louis, IL,
Northbrook, IL, Cincinnati, OH and Detroit, MI, respectively, with the highest concentrations
occurring in the summer. On average, the aromatic SOA concentrations made up 17 % of the
total SOA concentration. Thus, this work suggests that we are finding ambient PM levels on
an annual basis of about 0.15 ^tg/m3 associated with present toluene levels in the ambient air
in these Midwest cities. Based on preliminary analysis of recent laboratory experiments, it
appears the toluene tracer could also be formed during photooxidation of some of the
i 251
xylenes.
Over the past decade a variety of modeling studies have been conducted to predict
ambient SOA levels, with most studies focusing on the contributions of biogenic
monoterpenes and anthropogenic aromatic hydrocarbons. More recently, modelers have
begun to include the contribution of the isoprene SOA to ambient OC concentrations.252 In
general, the studies have been limited to comparing the sum of the POA and SOA
concentrations with ambient OC concentrations. The general consensus in the atmospheric
chemistry community appears to be that monoterpene contributions, which are clearly
significant, and the somewhat smaller aromatic contributions, are insufficient to account for
observed ambient SOA levels.253 Part of this gap has been filled recently by SOA predictions
for isoprene. Furthermore, the identification in ambient SOA of a tracer compound for the
sesquiterpene p-caryophyllene,254 coupled with the high sesquiterpene SOA yields measured
in the laboratory,255 suggests this class of hydrocarbons should be included in SOA chemical
mechanisms. In addition, recent data on SOA formation from aromatic hydrocarbons suggest
their contributions, while much smaller than biogenic hydrocarbons, could be larger than
previously thought.256'257
7.2.1.2.3 Ozone
As mentioned above, the addition of ethanol to fuels has been shown to contribute to
PAN formation and this is one way for it to contribute therefore to ground-level ozone
formation. PAN is a reservoir and carrier of NOx and is the product of acetyl radicals reacting
with NOi in the atmosphere. One source of PAN is the photooxidation of acetaldehyde
(Section 7.2.1.2.1), but any hydrocarbon having a methyl group has the potential for forming
acetyl radicals and therefore PAN.V PAN can undergo thermal decomposition with a lifetime
of approximately 1 hour at 298K or 148 days at 250K.w
CH3C(O)OONO2 + M -> CH3C(O)OO- + NO2 + M k = 3.3 x 10"4 s"1 258
v Many aromatic hydrocarbons, particularly those present in high percentages in gasoline (toluene, m-, o-, p-
xylene, and 1,3,5-, 1,2,4-trimethylbenzene), form methylglyoxal andbiacetyl, which are also strong generators
of acetyl radicals (Smith, D.F., T.E. Kleindienst, C.D. Mclver (1999) Primary product distribution from the
reaction of OH with m-, p-xylene and 1,2,4- and 1,3,5-Trimethylbenzene. J. Atmos. Chem., 34: 339- 364.).
w All rate coefficients in this RIA are listed at 298 K and, if applicable, 1 bar of air.
7-43
-------
Regulatory Impact Analysis
The reaction above shows how NC>2 is released in the thermal decomposition of PAN.
NOi can also be formed in photodegradation reactions where NO is converted to NO2 (see
OH radical reaction of acetaldehyde in Section 3.4.1.2.1). In both cases, NO2 further
photolyzes to produce ozone (Os).
NO2 + hv -> NO + O(3P) X = 300-800 nm 259
O(3P) + O2 + M -> O3 + M
The temperature sensitivity of PAN allows it to be stable enough at low temperatures
to be transported long distances before decomposing to release NO2. NO2 can then participate
in ozone formation in regions remote from the original NOx source.260 A discussion of CB05
mechanisms for ozone formation can be found in Yarwood et al. (2005).261
7.2.1.3 Modeling Uncertainties and Limitations
All the results presented below must be interpreted with the understanding that there
are uncertainties in inventories, atmospheric processes in CMAQ, and other aspects of the
modeling process. While it is beyond the scope of this Regulatory Impact Analysis to include
a comprehensive discussion of all limitations and uncertainties associated with air quality
modeling, several sources of uncertainty that impact analyses for this rule are addressed.
A key source of uncertainty is the photochemical mechanisms in CMAQ 4.7.
Pollutants such as ozone, PM, acetaldehyde, formaldehyde, acrolein, and 1,3-butadiene can be
formed secondarily through atmospheric chemical processes. Since secondarily formed
pollutants can result from many different reaction pathways, there are uncertainties associated
with each pathway. Simplifications of chemistry must be made in order to handle reactions of
thousands of chemicals in the atmosphere. Mechanisms for formation of ozone, PM,
acetaldehyde and peroxyacetyl nitrate (PAN) are discussed in Section 7.2.1.2.
For PM, there are a number of uncertainties associated with SOA formation that
should be addressed explicitly. As mentioned in Section 7.2.1.2.2, a large number of VOCs
emitted into the atmosphere, which have the potential to form SOA, have not yet been studied
in detail. In addition, the amount of ambient SOA that comes from benzene is uncertain.
Simplifications to the SOA treatment in CMAQ have also been made in order to preserve
computational efficiency. These simplifications are described in release notes for CMAQ 4.7
on the Community Modeling and Analysis System (CMAS) website.262
7.2.2 Air Quality Modeling Results
As described above, we performed a series of air quality modeling simulations for the
continental U.S in order to assess the impacts of the vehicle rule. We looked at impacts on
future ambient PM2.5, ozone, ethanol and air toxics levels, as well as nitrogen and sulfur
7-44
-------
Environmental and Health Impacts
deposition levels and visibility impairment. In this section, we present information on current
levels of pollution as well as model projected levels of pollution for 2030.
Emissions and air quality modeling decisions are made early in the analytical process.
For this reason, the inventories used in the air quality modeling and the benefits modeling,
which are presented in Section 5.8, are slightly different than the final vehicle standard
inventories presented in Section 5.5. However, the air quality inventories and the final rule
inventories are generally consistent, so the air quality modeling adequately reflects the effects
of the rule.
7.2.2.1 Particulate Matter (PM2.5 and PM10)
As described in Section 7.1, PM causes adverse health effects, and the EPA has set
national ambient air quality standards (NAAQS) to protect against those health effects. In this
section we present information on current and model-projected future PM levels.
7.2.2.1.1 Current Levels of PM
Figure 7-4 and Figure 7-5 show a snapshot of annual and 24-hour PMi.5
concentrations in 2008. There are two National Ambient Air Quality Standards (NAAQS) for
PM2.5: an annual standard (15 (^g/m3) and a 24-hour standard (35 (^g/m3). In 2008, the highest
annual average PM2.5 concentrations were in California, Arizona, and Hawaii and the highest
24-hour PM2.5 concentrations were in California and Virginia.
7-45
-------
Regulatory Impact Analysis
Annual
Concentration Range (pg/m3)
• 4.3-12.0 (610 Sites)
O 12.1 -15.0(221 Sites)
O 15.1 -18.0 (13 Sites)
• 18.1 - 23.5 (5 Sites)
Puerto Rico
Alaska
Figure 7-4 Annual Average PM2.s Concentrations in (ig/m3 for 2008
x From U.S. EPA, 2010. Our Nation's Air: Status and Trends through 2008. EPA-454/R-09-002. February
2010. Available at: http://www.epa.gov/airtrends/2010/index.html.
7-46
-------
Environmental and Health Impacts
24-hour
Concentration Range (|jg/m3)
• 8-15(51 Sites)
O 16-35 (743 Sites)
O 36-55 (43 Sites)
• 56-97 (12 Sites)
Puerto Rico
Alaska
Figure 7-5 24-hour (98th percentile 24- hour concentrations) PM2.s Concentrations in ng/m for 2008
The most recent revisions to the PM standards were in 1997 and 2006. In 2005, the
U.S. EPA designated nonattainment areas for the 1997 PM2.5 NAAQS (70 FR 19844, April
14, 2005).z As of January 6, 2010, approximately 88 million people live in the 39 areas that
are designated as nonattainment for the 1997 PM2.5 National Ambient Air Quality Standard
(NAAQS). These PM2.5 nonattainment areas are comprised of 208 full or partial counties.
Nonattainment areas for the 1997 PM2.5 NAAQS are pictured in Figure 7-6. On October 8,
2009, the EPA issued final nonattainment area designations for the 2006 24-hour PM2 5
NAAQS (74 FR 58688, November 13, 2009). These designations include 31 areas composed
Y From U.S. EPA, 2010. Our Nation's Air: Status and Trends through 2008. EPA-454/R-09-002. February
2010. Available at: http://www.epa.gov/airtrends/2010/index.html.
z 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.
7-47
-------
Regulatory Impact Analysis
of 120 full or partial counties with a population of over 70 million. Nonattainment areas for
the 2006 PMi.5 NAAQS are pictured in Figure 7-7. In total, there are 54 PM2.5 nonattainment
areas composed of 243 counties with a population of almost 102 million people.
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 will be required to
attain the 1997 PM2.5 NAAQS in the 2010 to 2015 time frame and then be required to
maintain the 1997 PM2.5 NAAQS thereafter.263 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
264
then be required to maintain the 2006 24-hour PM2 5 NAAQS thereafter. ' The vehicle
standards finalized here first apply to model year 2012 vehicles.
PM-2.5 Nonattainment Areas (1997 Standard)
Nonattainment areas are indicated by color.
When only a portion of a county is shown in color,
it indicates that only that part of the county is within
a nonattainment area boundary.
7C009
Figure 7-6 1997 PM2.5 Nonattainment Areas
7-48
-------
Environmental and Health Impacts
PM-2.5 Nonattainment Areas (2006 Standard)
Nonattainment areas are indicated by color.
When only a portion of a county is shown in color.
it indicates that only that part of the county is within
a nonattainment area boundary.
11/2009
Figure 7-7 2006 PM2.s Nonattainment Areas
As of January 6, 2010, approximately 26 million people live in the 47 areas that are
designated as nonattainment for the PMio NAAQS. There are 40 full or partial counties that
make up the PMio nonattainment areas. Nonattainment areas for the PMio NAAQS are
pictured in Figure 7-8.
7-49
-------
Regulatory Impact Analysis
Counties Designated N on attainment for PM-10
Newark Co., NY
(Moderate)
Mun. of Guaynabo, PR
(Moderate)
Juneau, AK
(Moderate)
Classification
BB Serious
I I Moderate
Classification colors are shown for whole counties and
denote the highest area classification that the county is in
Figure 7-8 PMio Nonattainment Areas
7.2.2.1.2 Projected Levels ofPM2.s
Generally, our modeling indicates that the vehicle standards will reduce PM2.5
concentrations in some localized areas of the country. In the following sections we describe
projected PM2.5 levels in the future, with and without the vehicle standards. Information on
the air quality modeling methodology is contained in Section 7.2.1.1. Additional detail can be
found in the air quality modeling technical support document (AQM TSD).
7-50
-------
Environmental and Health Impacts
7.2.2.1.2.1 Projected Levels of PM2.5 without this Rule
EPA has already adopted many mobile source emission control programs that are
expected to reduce ambient PM levels. These control programs include the New Marine
Compression-Ignition Engines at or Above 30 Liters per Cylinder rule,AA the Marine Spark-
Ignition and Small Spark-Ignition Engine rule (73 FR 59034, October 8, 2008), the
Locomotive and Marine Compression-Ignition Engine Rule (73 FR 25098, May 6, 2008), the
Clean Air Nonroad Diesel rule (69 FR 38957, June 29, 2004), the Heavy Duty Engine and
Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements (66 FR 5002, Jan.
18, 2001) and the Tier 2 Motor Vehicle Emissions Standards and Gasoline Sulfur Control
Requirements (65 FR 6698, Feb. 10, 2000). As a result of these and other federal, state and
local programs, the number of areas that fail to meet the PM^s NAAQS in the future is
expected to decrease. However, even with the implementation of all current state and federal
regulations, there are projected to be counties violating the PMi.5 NAAQS well into the
future.
The air quality modeling conducted projects that in 2030, with all current controls in
effect but excluding the emissions changes expected to occur as a result of the vehicle
standards being finalized here, at least 9 counties, with a projected population of nearly 28
million people, may not attain the annual standard of 15 (ig/m3 and at least 26 counties, with a
projected population of over 41 million people, may not attain the 2006 24-hour standard of
35 (^g/m3. Since the emission changes from this rule go into effect during the period when
some areas are still working to attain the PM2.5 NAAQS, the projected emission changes will
impact state and local agencies in their effort to attain and maintain the PM2.5 standard. In the
following section we discuss projected nonattainment areas and how they compare to the
areas which are projected to experience PM2.5 reductions or increases from the vehicle
standards.
7.2.2.1.2.2 Projected Annual Average PM2.s Design Values with this Rule
This section summarizes the results of our modeling of annual average PM^s air
quality impacts in the future due to the vehicle standards. Specifically, we compare a 2030
reference scenario, a scenario without the vehicle standards, to a 2030 control scenario which
includes the vehicle standards. Our modeling indicates that the majority of the modeled
counties will see decreases of less than 0.05 (Jg/m3 in their annual PM^s design values due to
AA This rule was signed on December 18, 2009 but has not yet been published in the Federal Register. The
signed version of the rule is available at http://epa.gov/otaq/oceanvessels.htm).
7-51
-------
Regulatory Impact Analysis
the vehicle standards. Figure 7-9 presents the changes in annual PM2.5 design values in
2030."
BB
Annual PM2.S DV Projections: 2030cpjdghg minus 2Q30cp
Figure 7-9 Projected Change in 2030 Annual PM2.5 Design Values Between the Reference Case and
Control Case
As shown in Figure 7-9, six counties will see decreases of more than 0.05
These counties are in southern California, central North Dakota, eastern Missouri, southwest
Louisiana and the Houston area in Texas. The maximum projected decrease in an annual
PMi.5 design value is 0.07 (^g/m3 in Harris County, Texas. The decreases in annual PMi.5
design values that we see in some counties are likely due to emission reductions related to
lower gasoline production at existing oil refineries; reductions in direct PMi.5 emissions and
PM2.5 precursor emissions (NOx and SOx) contribute to reductions in ambient concentrations
1 An annual PM2.s design value is the concentration that determines whether a monitoring site meets the annual
NAAQS for PM2 5. The full details involved in calculating an annual PM2 5 design value are given in appendix N
of 40 CFR part 50.
7-52
-------
Environmental and Health Impacts
of both direct PM2.5 and secondarily-formed PM2.5. Additional information on the upstream
emissions reductions that are projected with this final rule is available in Section 5.5.
There are also a few counties that will see small, no more than 0.01 (^g/m, design
value increases. These small increases in annual PM2.5 design values are likely related to
downstream emission increases. Additional information on the downstream emissions
increases that are projected with this final rule is also available in Section 5.3.3.5.
There are 9 counties, all in California, that are projected to have annual PM2.5 design
values above the NAAQS in 2030 with the vehicle standards in place. Table 7-2 below
presents the changes in design values for these counties.
Table 7-2 Change in Annual PM2.5 Design Values (ng/m3) for Counties Projected to be Above the Annual
PM2.5 NAAQS in 2030
County Name
Riverside Co., California
San Bernardino Co., California
Los Angeles Co., California
Kern Co., California
Tulare Co., California
Orange Co., California
Kings Co., California
Fresno Co., California
San Diego Co., California
Change in
Annual
PM2.5
Design
Value
-0.02
-0.03
-0.06
-0.01
-0.01
-0.03
0.00
0.00
0.00
Population
in 2030a
2,614,198
2,784,489
10,742,722
981,806
528,662
4,431,070
195,067
1,196,949
4,399,983
Note:
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. 2001.
Population by Single Year of Age CD.
Table 7-3 shows the average change in 2030 annual PM2.5 design values for: (1) all
counties with 2005 baseline design values, (2) counties with 2005 baseline design values that
exceeded the annual PM2.5 standard, (3) counties with 2005 baseline design values that did not
exceed the standard, but were within 10% of it, (4) counties with 2030 design values that
exceeded the annual PM2.5 standard, and (5) counties with 2030 design values that did not
exceed the standard, but were within 10% of it. Counties within 10% of the standard are
intended to reflect counties that although not violating the standards, will also be impacted by
changes in PM2 5 as they work to ensure long-term maintenance of the annual PM25 NAAQS.
These statistics show either no change or a decrease in annual PM2.5 design values in 2030.
7-53
-------
Regulatory Impact Analysis
On a population-weighted basis, the average modeled future-year annual PM2.5 design values
are projected to decrease by 0.01 (Jg/m3 due to the vehicle standards. On a population-
weighted basis annual PM2.5 design values in those counties that are projected to be above the
annual PM2.5 standard in 2030 will see a slightly larger decrease of 0.03 |^g/m3 due to the
vehicle standards.
Table 7-3 Average Change in Projected Annual PM2.5 Design Values
Average0
All
All, population-weighted
Counties whose 2005 base year is violating the
1997 annual PM2.5 standard
Counties whose 2005 base year is violating
the 1997 annual PM2.5 standard, population-
weighted
Counties whose 2005 base year is within 10
percent of the!997 annual PM2.5 standard
Counties whose 2005 base year is within 10
percent of the 1997 annual PM2.5 standard,
p opulation- weighted
Counties whose 2030 control case is violating
the 1997 annual PM2.5 standard
Counties whose 2030 control case is violating
the 1997 annual PM2.5 standard, population-
weighted
Counties whose 2030 control case is within 10%
of the 1997 annual PM2.5 standard
Counties whose 2030 control case is within 10%
of the 1997 annual PM2.5 standard, population-
weighted
Number
of US
Counties
576
70
102
9
5
2030
Population13
247,415,381
65,106,709
33,008,932
27,874,946
5,864,401
Change in
2030 design
value
(l^g/m3)
0.00
-0.01
0.00
-0.02
0.00
0.00
-0.02
-0.03
0.00
0.00
Note:
a Averages are over counties with 2005 modeled design values
b Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. 2001. Population by
Single Year of Age CD.
7.2.2.1.2.3 Projected 24-hour Average PM2.s Design Values with this Rule
This section summarizes the results of our modeling of 24-hour PM2.5 air quality
impacts in the future due to the vehicle standards. Specifically, we compare a 2030 reference
7-54
-------
Environmental and Health Impacts
scenario, a scenario without the vehicle standards, to a 2030 control scenario which includes
the vehicle standards. Our modeling indicates that the majority of the modeled counties will
see changes of between -0.05 (^g/m3 and +0.05 (^g/m3 in their 24-hour PM2.5 design values.
Figure 7-10 presents the changes in 24-hour PMi.5 design values in 2030.
.
cc
Difference In 24-hr PM2.5DVProjections: 2030cp_Mghg minus 2030cp
Figure 7-10 Projected Change in 2030 24-hour PM2.s Design Values Between the Reference Case and the
Control Case
As shown in Figure 7-10, 17 counties will see decreases of more than 0.05
These counties are in southern California, northern Utah, central North Dakota, eastern
Missouri, southern Arkansas, northern Oklahoma, southwest Louisiana and the Houston area
in Texas. The maximum projected decrease in a 24-hour PM2.5 design value is 0.21 |^g/m3 in
Harris County, Texas. The decreases in 24-hour PM2.5 design values that we see in some
counties are likely due to emission reductions related to lower gasoline production at existing
A 24-hour PM2.s design value is the concentration that determines whether a monitoring site meets the 24-
hour NAAQS for PM2 5. The full details involved in calculating a 24-hour PM2 5 design value are given in
appendix N of 40 CFR part 50.
7-55
-------
Regulatory Impact Analysis
oil refineries; reductions in direct PM2.5 emissions and PM2.5 precursor emissions (NOx and
SOx) contribute to reductions in ambient concentrations of both direct PMi.5 and secondarily-
formed PM2.5. Additional information on the upstream emissions reductions that are projected
with this final rule is available in Section 5.5.
There are also some counties that will see small, less than 0.05 (^g/m, design value
increases. These small increases in 24-hour PMi.5 design values are likely related to
downstream emissions increases.
There are 26 counties, mainly in California, that are projected to have 24-hour PMi.5
design values above the NAAQS in 2030 with the vehicle standards in place. Table 7-4
below presents the changes in design values for these counties.
7-56
-------
Environmental and Health Impacts
Table 7-4 Change in 24-hour
Design Values ((ig/m ) for Counties Projected to be Above the 24-hour
is NAAQS in 2030
County Name
Riverside Co., California
Kern Co., California
Allegheny Co., Pennsylvania
Fresno Co., California
San Bernardino Co., California
Los Angeles Co., California
Kings Co., California
Tulare Co., California
Lane Co., Oregon
Sacramento Co., California
Cache Co., Utah
Salt Lake Co., Utah
Orange Co., California
Butte Co., California
Stanislaus Co., California
Klamath Co., Oregon
Utah Co., Utah
Lincoln Co., Montana
Pierce Co., Washington
Santa Clara Co., California
Merced Co., California
Imperial Co., California
Wayne Co., Michigan
San Diego Co., California
Milwaukee Co., Wisconsin
Brooke Co., West Virginia
Change in 24-
hour PMi.5
Design Value
(l^g/m3)
-0.05
-0.02
0.04
-0.01
-0.02
-0.10
-0.01
0.00
0.00
-0.01
0.01
-0.02
-0.09
0.00
-0.01
0.00
0.04
0.00
0.03
-0.01
0.00
0.00
-0.01
-0.02
0.00
0.00
Population in
2030a
2,614,198
981,806
1,234,931
1,196,950
2,784,490
10,742,722
195,067
528,663
460,993
1,856,971
141,446
1,431,946
4,431,071
287,236
688,246
77,200
661,456
20,454
1,082,579
2,320,199
313,334
174,175
1,838,270
4,399,983
927,986
24,095
Note:
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. 2001.
Population by Single Year of Age CD.
Table 7-5 shows the average change in 2030 24-hour PM2.5 design values for: (1) all
counties with 2005 baseline design values, (2) counties with 2005 baseline design values that
exceeded the 24-hour PM2.5 standard, (3) counties with 2005 baseline design values that did
not exceed the standard, but were within 10% of it, (4) counties with 2030 design values that
7-57
-------
Regulatory Impact Analysis
exceeded the 24-hour PM2.5 standard, and (5) counties with 2030 design values that did not
exceed the standard, but were within 10% of it. Counties within 10% of the standard are
intended to reflect counties that although not violating the standards, will also be impacted by
changes in PM2.5 as they work to ensure long-term maintenance of the 24-hour PM2.5
NAAQS. On a population-weighted basis, the average modeled future-year 24-hour PM2.5
design values are projected to decrease by 0.01 (Jg/m3 due to the vehicle standards. On a
population- weighted basis 24-hour PM2.5 design values in those counties that are projected to
be above the 24-hour PM2.5 standard in 2030 will see a slightly larger decrease of 0.05 |^g/m3.
Table 7-5 Average Change in Projected 24-hour
Design Values
Average0
All
All, population-weighted
Counties whose 2005 base year is violating the
2006 24-hour PM2.5 standard
Counties whose 2005 base year is violating the
2006 24-hour PM2.5 standard, population-
weighted
Counties whose 2005 base year is within 10
percent of the 2006 24-hour PM2.5 standard
Counties whose 2005 base year is within 10
percent of the 2006 24-hour PM2.5 standard,
p opulation- weighted
Counties whose 2030 control case is violating
the 2006 24-hour PM2.5 standard
Counties whose 2030 control case is violating
the 2006 24-hour PM2.5 standard, population-
weighted
Counties whose 2030 control case is within 10%
of the 2006 24-hour PM2.5 standard
Counties whose 2030 control case is within 10%
of the 2006 24-hour PM2.5 standard,
p opulation- weighted
Number
of US
Counties
579
105
139
26
24
2030
Population13
247,228,608
86,013,770
53,848,276
41,416,465
18,526,165
Change in
2030 design
value
(l^g/m3)
0.00
-0.01
0.00
-0.02
0.00
0.00
-0.01
-0.05
0.00
0.01
Note:
a Averages are over counties with 2005 modeled design values
b Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. 2001. Population by
Single Year of Age CD.
7-58
-------
Environmental and Health Impacts
7.2.2.2 Ozone
As described in Section 7.1, ozone causes adverse health effects, and the EPA has set
national ambient air quality standards (NAAQS) to protect against those health effects. In this
section, we present information on current and model-projected future ozone levels.
7.2.2.2.1 Current Levels of Ozone
Figure 7-11 shows a snapshot of ozone concentrations in 2007. The highest ozone
concentrations were located in California. Thirty-two percent of the sites were above 0.075
ppm, the level of the 2008 standard.
Concentration Range (ppm)
• 0.029 - 0.059 (89 Sites)
O 0.060 - 0.075 (722 Sites)
O 0.076 - 0.095 (336 Sites)
• 0.096-0.120(41 Sites)
Puerto Rico
Alaska
Figure 7-11 Ozone Concentrations (fourth highest daily maximum 8-hour concentration) in ppm for
2008DD
DD From U.S. EPA, 2010. Our Nation's Air: Status and Trends through 2008. EPA-454/R-09-002. February
2010. Available at: http://www.epa.gov/airtrends/2010/index.html.
7-59
-------
Regulatory Impact Analysis
Nonattainment and Maintenance Areas in the U .3.
8-hour Ozone (1997 Standard)
1/2010
[Nonattainment Areas (238 entire counties)
Nonattainment Areas (28 partial counties)
Maintenance Areas (168 entire or partial counties)
Partial counties, those with part of the county designated
nonattainment and part attainment, are shown as full counties on this map.
Figure 7-12 1997 Ozone Nonattainment Areas
The primary and secondary national ambient air quality standards (NAAQS) for ozone
are 8-hour standards set at 0.075 ppm. The most recent revision to the ozone standards was in
2008; the previous 8-hour ozone standards, set in 1997, had been set at 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). As of January 6, 2010, there are 51 8-hour ozone nonattainment areas
for the 1997 ozone NAAQS composed of 266 full or partial counties with a total population
of over 122 million. Figure 7-12 presents the 1997 NAAQS ozone nonattainment areas. On
January 6, 2010, EPA proposed to reconsider the 2008 ozone NAAQS to ensure that they are
requisite to protect public health with an ample margin of safety, and requisite to protect
public welfare. EPA intends to complete the reconsideration by August 31, 2010. If, as a
result of the reconsideration, EPA promulgates different ozone standards, the new 2010 ozone
7-60
-------
Environmental and Health Impacts
standards would replace the 2008 ozone standards and the requirement to designate areas for
the replaced 2008 standards would no longer apply. Because of the significant uncertainty the
reconsideration proposal creates regarding the continued applicability of the 2008 ozone
NAAQS, EPA has extended the deadline for designating areas for the 2008 NAAQS by one
year. This will allow EPA to complete its reconsideration of the 2008 ozone NAAQS before
determining whether designations for those standards are necessary.
If EPA promulgates new ozone standards in 2010, EPA intends to accelerate the
designations process for the primary standard so that the designations would be effective in
August 2011. EPA is considering two alternative schedules for designating areas for a new
seasonal secondary standard, an accelerated schedule or a 2-year schedule.
Table 7-6 includes an estimate, based on 2006-08 air quality data, of the counties with
design values greater than the 2008 ozone NAAQS.
Table 7-6 Counties with Design Values Greater Than the Ozone NAAQS
1997 Ozone Standard: counties within the 54
areas currently designated as nonattainment (as
of 1/6/10)
2008 Ozone Standard: additional counties that
would not meet the 2008 NAAQS (based on
2006-2008 air quality data)b
Total
NUMBER OF
COUNTIES
266
156
422
POPULATIONA
122,343, 799
36,678,478
159,022,277
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.
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 be required to maintain
it thereafter.EE In addition, there will be attainment dates associated with the designation of
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 their attainment date June 15, 2024. The San Joaquin Valley Air Basin 8-
7-61
-------
Regulatory Impact Analysis
nonattainment areas as a result of the reconsideration of the 2008 ozone NAAQS. If the
ozone NAAQS reconsideration action is completed on the proposed schedule, the primary
NAAQS attainment dates would be in the 2014-2031 time frame. The vehicle standards first
apply to model year 2012 vehicles.
7.2.2.2.2 Projected Levels of Ozone
In the following sections, we describe projected ozone levels in the future with and
without the vehicle standards. We do not expect this rule to have a meaningful impact on
ozone concentrations, given the small magnitude of the ozone impacts and the fact that much
of the impact is due to ethanol assumptions that are independent of this rule. Our modeling
indicates that there will be increases in ozone design value concentrations in many areas of
the country and decreases in ozone design value concentrations in a few areas. However, the
increases in ozone design values are not due to the standards finalized in this rule, but are
related to our assumptions about the volume of ethanol that will be blended into gasoline.
The ethanol volumes will be occurring as a result of the recent Renewable Fuel Standards
(RFS2) rule.265 Information on the air quality modeling methodology is contained in Section
7.2.1.1. Additional detail can be found in the air quality modeling technical support
document (AQM TSD).
7.2.2.2.2.1 Projected Levels of Ozone without this Rule
EPA has already adopted many emission control programs that are expected to reduce
ambient ozone levels. These control programs include the New Marine Compression-Ignition
Engines at or Above 30 Liters per Cylinder rule,FF the Marine Spark-Ignition and Small
Spark-Ignition Engine rule (73 FR 59034, October 8, 2008), the Locomotive and Marine Rule
(73 FR 25098, May 6, 2008), the Clean Air Interstate Rule (70 FR 25162, May 12, 2005), the
Clean Air Nonroad Diesel rule (69 FR 38957, June 29, 2004), and the Heavy Duty Engine and
Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements (66 FR 5002, Jan.
18, 2001). As a result of these and other federal, state and local programs, 8-hour ozone
levels are expected to improve in the future. However, even with the implementation of all
current state and federal regulations, there are projected to be counties violating the ozone
NAAQS well into the future.
The air quality modeling projects that in 2030, with all current controls in effect but
excluding the emissions changes expected to occur as a result of this final rule, at least 16
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
their attainment date June 15, 2024.
FF This rule was signed on December 18, 2009 but has not yet been published in the Federal Register. The
signed version of the rule is available at http://epa.gov/otaq/oceanvessels.htm).
7-62
-------
Environmental and Health Impacts
counties, with a projected population of almost 35 million people, may not attain the 2008 8-
hour ozone standard of 75 ppb. Since the emission changes from this rule go into effect
during the period when some areas are still working to attain the ozone NAAQS, the projected
emission changes will impact state and local agencies in their effort to attain and maintain the
ozone standard. In the following section we discuss projected nonattainment areas and how
they compare to the areas which are projected to experience ozone reductions from the
vehicle standards.
7.2.2.2.2.2 Projected Levels of Ozone with this Rule
This section summarizes the results of our modeling of ozone air quality impacts in
the future with the vehicle standards. Specifically, we compare a 2030 reference scenario, a
scenario without the vehicle standards, to a 2030 control scenario which includes the vehicle
standards. We do not expect this rule to have a meaningful impact on ozone concentrations,
given the small magnitude of the ozone impacts and the fact that much of the impact is due to
ethanol assumptions that are independent of this rule.
Our modeling indicates ozone design value concentrations will increase in many areas
of the country and decrease in a few areas. The increases in ozone design values are not due
to the standards finalized in this rule, but are related to our assumptions about the volume of
ethanol that will be blended into gasoline. The ethanol volumes will be occurring as a result
of the recent RFS2 rule. As discussed in Sections 5.3.2 and 5.3.3.5 of this RIA, we attribute
decreased fuel consumption and production from this program to gasoline only, while
assuming constant ethanol volumes in our reference and control cases. Holding ethanol
volumes constant while decreasing gasoline volumes increases the market share of 10%
ethanol (E10) in the control case. However, the increased E10 market share is projected to
occur regardless of this rule, and the air quality impacts of this effect are included in our
analyses for the recent RFS2 rule. As the RFS2 analyses indicate, increasing usage of E10
fuels (when compared with EO fuels) can increase NOx emissions and thereby increase ozone
concentrations, especially in NOx-limited areas where relatively small amounts of NOx
enable ozone to form rapidly.266 Figure 7-13 presents the changes in 8-hour ozone design
value concentration in 2030 between the reference case and the control case.00
00 An 8-hour ozone design value is the concentration that determines whether a monitoring site meets the 8-hour
ozone NAAQS. The full details involved in calculating an 8-hour ozone design value are given in appendix I of
40 CFR part 50.
7-63
-------
Regulatory Impact Analysis
Difference in 8-hour Ozone DV: 2030cp_ldghg minus 2030cp
Figure 7-13 Projected Change in 2030 8-hour Ozone Design Values Between the Reference Case and
Control Case
As can be seen in Figure 7-13, the majority of the design value increases are less than
0.1 ppb. However, there are some counties that will see 8-hour ozone design value increases
above 0.1 ppb; these counties are along the mid-Atlantic coast and in southern Arizona. The
maximum projected increase in an 8-hour ozone design value is 0.25 ppb in Richland County,
South Carolina. There are also some counties that are projected to see 8-hour ozone design
value decreases. The decreases in ambient ozone concentration are likely due to projected
upstream emissions decreases in NOx and VOCs from reduced gasoline production. The
counties with ozone design value decreases greater than 0.1 ppb are in California, Texas,
Louisiana, Mississippi, Kentucky, Ohio and West Virginia. The maximum decrease projected
in an 8-hour ozone design value is 0.22 ppb in Riverside, CA.
There are 16 counties, half of them in California, that are projected to have 8-hour
ozone design values above the 2008 NAAQS in 2030 with the vehicle standards in place.
Table 7-7 below presents the changes in design values for these counties. Increases in design
values in Maryland and Connecticut are a reflection of our ethanol volume assumptions as
7-64
-------
Environmental and Health Impacts
discussed above (also Sections 5.3.2 and 5.3.3.5) and are not due to the standards finalized in
this rule.
Table 7-7 Change in Ozone Design Values (ppb) for Counties Projected to be Above the 2008 Ozone
NAAQS in 2030
County Name
San Bernardino Co.,
California
Riverside Co., California
Los Angeles Co., California
Kern Co., California
Harris Co., Texas
Tulare Co., California
Suffolk Co., New York
Fresno Co., California
Brazoria Co., Texas
Orange Co., California
Harford Co., Maryland
Fairfield Co., Connecticut
East Baton Rouge Co.,
Louisiana
Calaveras Co., California
Ventura Co., California
New Haven Co., Connecticut
San Bernardino Co.,
California
Change in
8-hour
Ozone
Design
Value
(ppb)
-0.18
-0.22
-0.06
-0.07
-0.17
-0.02
-0.02
-0.04
-0.15
-0.13
0.06
0.00
-0.10
-0.10
-0.03
0.04
-0.18
Population
in 2030a
60,710,005
60,658,001
60,376,012
60,295,001
482,010,05
5
61,070,009
361,030,00
2
60,190,008
480,391,00
4
60,595,001
240,251,00
1
90,013,007
220,330,00
3
60,090,001
61,112,002
90,093,002
60,710,005
Note:
a Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. 2001.
Population by Single Year of Age CD.
Table 7-8 shows the average change in 2030 8-hour ozone design values for: (1) all
counties with 2005 baseline design values, (2) counties with 2005 baseline design values that
exceeded the 2008 ozone standard, (3) counties with 2005 baseline design values that did not
7-65
-------
Regulatory Impact Analysis
exceed the 2008 standard, but were within 10% of it, (4) counties with 2030 design values
that exceeded the 2008 ozone standard, and (5) counties with 2030 design values that did not
exceed the standard, but were within 10% of it. Counties within 10% of the standard are
intended to reflect counties that although not violating the standards, will also be impacted by
changes in ozone as they work to ensure long-term maintenance of the ozone NAAQS. Many
of these statistics, except for counties with 2030 design values that exceed the 2008 ozone
standard, show an increase in 2030. Again, increases in ozone design value concentrations
are a reflection of our ethanol volume assumptions, as discussed above (also Sections 5.3.2
and 5.3.3.5), and are not due to the standards finalized in this rule. On a population-weighted
basis, the average modeled future-year 8-hour ozone design values are projected to increase
by 0.28 ppb in 2030. On a population-weighted basis those counties that are projected to be
above the 2008 ozone standard in 2030 will see a decrease of 0.10 ppb due to the vehicle
standards.
Table 7-8 Average Change in Projected 8-hour Ozone Design Value
Average0
All
All, population-weighted
Counties whose 2005 base year is violating the
2008 8-hour ozone standard
Counties whose 2005 base year is violating the
2008 8-hour ozone standard, population-
weighted
Counties whose 2005 base year is within 10
percent of the 2008 8-hour ozone standard
Counties whose 2005 base year is within 10
percent of the 2008 8-hour ozone standard,
p opulation- weighted
Counties whose 2030 control case is violating
the 2008 8-hour ozone standard
Counties whose 2030 control case is violating
the 2008 8-hour ozone standard, population-
weighted
Counties whose 2030 control case is within 10%
of the 2008 8-hour ozone standard
Counties whose 2030 control case is within 10%
of the 2008 8-hour ozone standard, population-
weighted
Number
of US
Counties
678
389
208
16
80
2030
Population13
262,264,195
192,026,888
47,276,756
34,751,421
61,467,398
Change in
2030 design
value (ppb)
0.03
0.01
0.03
0.01
0.03
0.02
-0.07
-0.10
0.00
0.01
Note:
a Averages are over counties with 2005 modeled design values
7-66
-------
Environmental and Health Impacts
b Population numbers based on Woods & Poole data. Woods & Poole Economics, Inc. 2001. Population by
Single Year of Age CD.
Ground-level ozone pollution is formed by the reaction of VOCs and NOx in the
atmosphere in the presence of heat and sunlight. The science of ozone formation, transport,
and accumulation is complex.267 The projected ozone decreases which are seen in the air
quality modeling for this final rule are likely a result of the emissions changes due to the
vehicle standards combined with the photochemistry involved, the different background
concentrations of VOCs and NOx in different areas of the country, and the different
meteorological conditions in different areas of the country.
When VOC levels are relatively high, relatively small amounts of NOx enable ozone
to form rapidly. Under these conditions, VOC reductions have little effect on ozone and
while NOx reductions are highly effective in reducing ozone, NOx increases lead to increases
in ozone. 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. Rural areas are usually NOx -limited, due to the relatively large amounts of biogenic
VOC emissions in such areas.
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.
7.2.2.3 Air Toxics
7.2.2.3.1 Current Levels of Air Toxics
The majority of Americans continue to be exposed to ambient concentrations of air
toxics at levels which have the potential to cause adverse health effects.268 The levels of air
toxics to which people are exposed vary depending on where people live and work and the
kinds of activities in which they engage, as discussed in detail in U.S. EPA's most recent
9^0
Mobile Source Air Toxics Rule. In order to identify and prioritize air toxics, emission
source types and locations which are of greatest potential concern, U. S. EPA conducts the
National-Scale Air Toxics Assessment (NATA). The most recent NATA was conducted for
calendar year 2002, and was released in June 2009.27° NATA for 2002 includes four steps:
1) Compiling a national emissions inventory of air toxics emissions from outdoor
sources
2) Estimating ambient concentrations of air toxics across the United States
3) Estimating population exposures across the United States
7-67
-------
Regulatory Impact Analysis
4) Characterizing potential public health risk due to inhalation of air toxics including
both cancer and noncancer effects
Figure 7-14 and Figure 7-15 depict estimated county-level carcinogenic risk and
noncancer respiratory hazard from the assessment. The respiratory hazard is dominated by a
single pollutant, acrolein.
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. 271'HH Benzene is the largest contributor to cancer risk of all 124 pollutants
quantitatively assessed in the 2002 NATA, and mobile sources were responsible for 59
percent of benzene emissions in 2002. Over the years, EPA has implemented a number of
mobile source and fuel controls which have resulted in VOC reductions, which also reduced
benzene and other air toxic emissions.
2002 National Scale Assessment
Estimated County Level Carcinogenic Risk
«1 in a Milan
1 - 25 In a Milton
Z5 - 50 in a Milton
50 - 75 in a ! hi '.
75-100. in a Milan
>100 In a Mlfton
Figure 7-14 County Level Average Carcinogenic Risk, 2002 NATA
' NATA relies on a Guassian plume model, Assessment System for Population Exposure Nationwide
(ASPEN), to estimate toxic air pollutant concentrations. Projected air toxics concentrations presented in this rule
were modeled with CMAQ 4.7, which has only recently been updated to include air toxics.
7-68
-------
Environmental and Health Impacts
2002 National Scale Assessment
Estimated County Level Noncancer (Respiratory) Risk
Figure 7-15 County Level Average Noncancer Hazard Index, 2002 NATA
7.2.2.3.2 Projected Levels of Air Toxics
In the following sections, we describe results of our modeling of air toxics levels in
the future with the standards finalized in this action. Although there are a large number of
compounds which are considered air toxics, we focused on those which were identified as
national and regional-scale cancer and noncancer risk drivers in past NATA assessments and
were also likely to be significantly impacted by the standards. These compounds include
benzene, 1,3-butadiene, formaldehyde, acetaldehyde, and acrolein. Ethanol impacts were also
included in our analyses. Information on the air quality modeling methodology is contained
in Section 7.2.1. Additional detail such as the seasonal concentration maps for the modeled
air toxics can be found in the air quality modeling technical support document (AQM TSD) in
the docket for this rule.
It should be noted that EPA has adopted many mobile source emission control
programs that are expected to reduce ambient air toxics levels. These control programs
include the Heavy-duty Onboard Diagnostic Rule (74 FR 8310, February 24, 2009), Small SI
and Marine SI Engine Rule (73 FR 59034, October 8, 2008), Locomotive and Commercial
Marine Rule (73 FR 25098, May 6, 2008), Mobile Source Air Toxics Rule (72 FR 8428,
February 26, 2007), Clean Air Nonroad Diesel Rule (69 FR 38957, June 29, 2004), Heavy
7-69
-------
Regulatory Impact Analysis
Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements
(66 FR 5002, Jan. 18, 2001) and the Tier 2 Motor Vehicle Emissions Standards and Gasoline
Sulfur Control Requirements (65 FR 6698, Feb. 10, 2000). As a result of these programs, the
ambient concentration of air toxics in the future is expected to decrease. The reference case
and control case scenarios include these controls.
Our modeling indicates that the GHG standards have relatively little impact on
national average ambient concentrations of the modeled air toxics. Because overall impacts
are small, we concluded that assessing exposure to ambient concentrations and conducting a
quantitative risk assessment of air toxic impacts was not warranted. However, we did develop
population metrics, including the population living in areas with increases or decreases in
concentrations of various magnitudes.
Acetaldehyde
Our air quality modeling does not show substantial overall nationwide impacts on
ambient concentrations of acetaldehyde as a result of the standards finalized in this rule.
Annual percent changes in ambient concentrations of acetaldehyde are less than 1% across the
country (Figure 7-16). Decreases in ambient concentrations of acetaldehyde seen in the much
of the eastern half of the U.S. and parts of the West are generally less than 0.01 |^g/m3 (Figure
7-16).
, far Acealdehyde
Figure 7-16 Changes in Acetaldehyde Ambient Concentrations Between the Reference Case and the
Control Case in 2030: Percent Changes (left) and Absolute Changes in ng/m3 (right)
7-70
-------
Environmental and Health Impacts
Formaldehyde
Our modeling projects that the standards finalized in this rule will not have a
significant impact on ambient formaldehyde concentrations. As shown in Figure 7-17, annual
percent changes in ambient concentrations of formaldehyde are less than 1% across the
country, with the exception of a 1 to 5% decrease in a small area of southern Kansas and
northern Oklahoma. Figure 7-17 also shows that absolute changes in ambient concentrations
of formaldehyde are generally less than 0.1 |^g/m3. Increases in ambient formaldehyde
concentrations, which range from 0.001 to 0.1 (^g/m3, are a reflection of our ethanol volume
assumptions as discussed above in Section 7.2.2.2.2.2 (also Sections 5.3.2 and 5.3.3.5) and are
not due to the standards finalized in this rule.
Map eotors do not into*!* lh« svwity of i*r
Sc*i« rang*! and incmrrnnts may net b*
Change • 20JOcpJdgHg minus 20Mcp, Soi Four,-!', '••; . ..'>-
Figure 7-17 Changes in Formaldehyde Ambient Concentrations Between the Reference Case and the
Control Case in 2030: Percent Changes (left) and Absolute Changes in ng/m3 (right)
Ethanol
Our air quality modeling results do not show substantial impacts on ambient
concentrations of ethanol from the vehicle GHG standards. While Figure 7-18 shows
increases in ambient ethanol concentrations ranging between 1 and 50% in some areas of the
country, these increases are a reflection of our ethanol volume assumptions as discussed
7-71
-------
Regulatory Impact Analysis
above in Section 7.2.2.2.2.2 (also Sections 5.3.2 and 5.3.3.5) and are not due to the standards
finalized in this rule.
' ..-:•- •:•',: -..-•;<• x<._,!-,y> .:•..•• r-.-Y - p.- ,
Figure 7-18 Changes in Ethanol Ambient Concentrations Between the Reference Case and the Control
Case in 2030: Percent Changes (left) and Absolute Changes in (ig/m3 (right)
Benzene
Our modeling projects that the standards finalized in this rule will not have a
significant impact on ambient benzene concentrations. Figure 7-19 shows decreases in
ambient benzene concentrations ranging between 1 and 10% and between 0.001 and 0.1
l^g/m3. Because this rule will reduce consumption and production of gasoline, some of these
decreases in benzene concentrations are likely due to the vehicle GHG standards. However,
decreases in benzene concentrations may also be a reflection of our ethanol volume
assumptions as discussed above for ozone, ethanol and formaldehyde, and are not due to the
standards finalized in this rule. For example, the percent change map in Figure 7-19 below
shows benzene decreases occurring in the same areas of the country as ozone, ethanol, and
formaldehyde increases.
7-72
-------
Environmental and Health Impacts
2030cp. farBsraene
Figure 7-19 Changes in Benzene Ambient Concentrations Between the Reference Case and the Control
Case in 2030: Percent Changes (left) and Absolute Changes in ng/m3 (right)
1,3-Butadiene
Our air quality modeling results do not show substantial impacts on ambient
concentrations of 1,3-butadiene from the GHG standards. Small decreases ranging from 1 to
10% occur in some southern areas of the country and increases ranging from 1 to over 100%
occur in some northern areas and areas with high altitudes (Figure 7-20). Changes in absolute
concentrations of ambient 1,3-butadiene are less than 0.001 |^g/m3 except in some areas of the
Northeast and Utah (Figure 7-20). Annual increases in ambient concentrations of 1,3-
butadiene are driven by wintertime rather than summertime changes (seasonal maps can be
found in the AQM TSD). These increases appear in rural areas with cold winters and low
ambient levels but high contributions of emissions from snowmobiles, and a major reason for
this modeled increase may be deficiencies in available emissions test data used to estimate
snowmobile 1,3-butadiene emission inventories. These data were based on tests using only
three engines, which showed significantly higher 1,3-butadiene emissions with 10% ethanol.
However, they may not have been representative of real-world response of snowmobile
engines to ethanol. Regardless, these increases are a reflection of our ethanol volume
assumptions and are not due to the standards finalized in this rule.
7-73
-------
Regulatory Impact Analysis
2030cp. for 1.3- Birtaatem?
Figure 7-20 Changes in 1,3-Butadiene Ambient Concentrations Between the Reference Case and the
Control Case in 2030: Percent Changes (left) and Absolute Changes in ng/m3 (right)
Acrolein
Our air quality modeling results do not show substantial impacts on ambient
concentrations of acrolein from the standards finalized in this rule. Small decreases ranging
from 1 to 2.5% occur in a few areas of the country and increases ranging from 1 to 100%
occur in some northern areas and areas with high altitudes (Figure 7-21). Changes in absolute
concentrations of acrolein are less than 0.001 |^g/m3 across the country (Figure 7-21).
Ambient acrolein increases are driven by wintertime changes (see the AQM TSD for seasonal
maps), and occur in the same areas of the country that have wintertime rather than
summertime increases in ambient 1,3-butadiene. 1,3-butadiene is a precursor to acrolein, and
these increases are likely associated with the same emission inventory uncertainties in areas of
high snowmobile usage seen for 1,3-butadiene. As described above, these increases are a
reflection of our ethanol volume assumptions and are not due to the standards finalized in this
rule.
7-74
-------
Environmental and Health Impacts
2030cp, torAcraiein
2036tp. for Acrolein
Figure 7-21 Changes in Acrolein Ambient Concentrations Between the Reference Case and the Control
Case in 2030: Percent Changes (left) and Absolute Changes in ng/m3 (right)
Population Metrics
To assess the impact of projected changes in air quality with the GHG standards, we
developed population metrics that show population experiencing increases and decreases in
annual ambient concentrations across the modeled air toxics. Table 7-9 illustrates the
percentage of the population impacted by changes of various magnitudes in annual ambient
concentrations between the reference case and the control case. As discussed above, increases
in ambient ethanol, 1,3-butadiene, and acrolein concentration are due to our ethanol volume
assumptions, and are not the result of the GHG vehicle standards.
7-75
-------
Regulatory Impact Analysis
Table 7-9 Percent of Total Population Impacted by Changes in Annual Ambient Concentrations of Toxic
Pollutants Between the Reference and Control Cases in 2030
Percent Change
in Annual
Ambient
Concentration
<-100
>-100to<-50
>-50to<-10
>-10to<-5
>-5 to <-2.5
>-2.5 to <-l
> -1 to <1
>1 to <2.5
>2.5 to<5
>5 to <10
>10 to <50
>50 to <100
>100
Acetaldehyde
0.61%
5.13%
94.86%
Acrolein
0.97%
0.18%
1.92%
94.82%
2.19%
0.93%
0.14%
0.63%
Benzene
0.60%
4.65%
16.22%
78.63%
1,3-Butadiene
0.64%
3.26%
8.43%
75.36%
8.45%
2.25%
0.82%
0.78%
0.27%
0.13%
Ethanol
0.44%
1.84%
15.11%
67.59%
4.92%
2.77%
3.96%
3.82%
Formaldehyde
0.54%
0.41%
0.27%
99.75%
7.2.2.4 Deposition of Nitrogen and Sulfur
7.2.2.4.1 Current Levels of Nitrogen and Sulfur Deposition
Over the past two decades, the EPA has undertaken numerous efforts to reduce
nitrogen and sulfur deposition across the U.S. Analyses of long-term monitoring data for the
U.S. show that deposition of both nitrogen and sulfur compounds has decreased over the last
17 years although many areas continue to be negatively impacted by deposition. Deposition
of inorganic nitrogen and sulfur species routinely measured in the U.S. between 2004 and
2006 were as high as 9.6 kilograms of nitrogen per hectare per year (kg N/ha/yr) and 21.3
kilograms of sulfur per hectare per year (kg S/ha/yr). Figure 7-22 and Figure 7-23 show that
annual total deposition (the sum of wet and dry deposition) decreased between 1989-1999 and
2004-2006 due to sulfur and NOx controls on power plants, motor vehicles and fuels in the
U.S. The data show that reductions were more substantial for sulfur compounds than for
nitrogen compounds. These numbers are generated by the U.S. national monitoring network
and they likely underestimate nitrogen deposition because neither ammonia nor organic
nitrogen is measured. In the eastern U.S., where data are most abundant, total sulfur
deposition decreased by about 44% between 1990 and 2007, while total nitrogen deposition
decreased by 25% over the same time frame.272
7-76
-------
Environmental and Health Impacts
••
* Annjilatti ullur«ip«ltloii. 1WS-19S1
13
131
126 11J
C c
„
i:* '*"'
IB'
'f imiiji r
sa 73
i tu linn mi
L'i
—
-20
10 ::<,'• ' •
i I p«
»iiimtin -iinirinp-iirr
cf tool sKBr *K*Bc«t
Figure 7-22 Total Sulfur Deposition in the Contiguous U.S., 1989-1991 and 2004 -2006
7-77
-------
Regulatory Impact Analysis
A Ming* rani* nrugtii dtp melon
u
4
OS
I".
C
—IS,
rflcat* toui nmoya i
Inn (UayanafH
0 &to ddr&n b&ut* to* nUttnt'jffitu* aid* stKiipf-
Cotan h eMH Mm ffw brukdoiM ef
Dry nitrogen
-------
Environmental and Health Impacts
7.2.2.4.2 Projected Levels of Nitrogen and Sulfur Deposition
Our air quality modeling does not show substantial overall nationwide impacts on the
annual total sulfur and nitrogen deposition occurring across the U.S. as a result of the vehicle
standards required by this rule. Figure 7-24 shows that for sulfur deposition the vehicle
standards will result in annual percent decreases of 0.5% to more than 2% in locations with
refineries as a result of the lower output from refineries due to less gasoline usage. These
locations include the Texas and Louisiana portions of the Gulf Coast; the Washington D.C.
area; Chicago, IL; portions of Oklahoma and northern Texas; Bismarck, North Dakota;
Billings, Montana; Casper, Wyoming; Salt Lake City, Utah; Seattle, Washington; and San
Francisco, Los Angeles, and San Luis Obispo, California. The remainder of the country will
see only minimal changes in sulfur deposition, ranging from decreases of less than 0.5% to
increases of less than 0.5%. The impacts of the vehicle standards on nitrogen deposition are
minimal, ranging from decreases of up to 0.5% to increases of up to 0.5%.
V
'inAimual Tata! Sulfur Deposition {2Q30cp_letytig minus WSOcp)
Figure 7-24 Percent Change in Annual Total Sulfur over the U.S. Modeling Domain as a Result of the
Required Vehicle Standards
7-79
-------
Regulatory Impact Analysis
7.2.2.5 Visibility Degradation
7.2.2.5.1 Current Visibility Levels
Recently designated PM2.5 nonattainment areas indicate that, as of January 6, 2010,
approximately 101 million people live in nonattainment areas for the PMi.5 NAAQS. Thus, at
least these populations would likely be experiencing visibility impairment, as well as many
thousands of individuals who travel to these areas. In addition, while visibility trends have
improved in mandatory class I federal areas, the most recent data show that these areas
continue to suffer from visibility impairment. In eastern areas, average visual range has
decreased from 90 miles to 15-25 miles. In western areas, visual range has decreased from
•"1"7Q
140 miles to 35-90 miles. In summary, visibility impairment is experienced throughout the
U.S., in multi-state regions, urban areas, and remote mandatory class I federal areas.
7.2.2.5.2 Projected Visibility Levels
Air quality modeling conducted for this final rule was used to project visibility
conditions in 138 mandatory class I federal areas across the U.S. in 2030. The results show
that all the modeled areas will continue to have annual average deciview levels above
background in 2030.11 The results also indicate that the majority of the modeled mandatory
class I federal areas will see no change in their visibility, but some mandatory class I federal
areas will see improvements in visibility due to the vehicle standards and a few mandatory
class I federal areas will see visibility decreases. The average visibility at all modeled
mandatory class I federal areas on the 20% worst days is projected to improve by 0.002
deciviews, or 0.01%, in 2030. The greatest improvement in visibilities will be seen in Bosque
de Apache (New Mexico) and the San Gorgonio Wilderness (near Los Angeles, California).
Bosque de Apache will see a 0.15% improvement (0.02 DV) and the San Gorgonio
Wilderness will see a 0.10% improvement (0.02 DV) in 2030 due to the vehicle standards.
The following six areas will see a degradation of 0.01 DV in 2030 as a result of the vehicle
standards: Hells Canyon Wilderness (Oregon), 0.06% degradation; Kalmiopsis Wilderness
(Oregon), 0.06% degradation; Strawberry Mountain Wilderness (Oregon), 0.06%
degradation; Petrified Forest National Park (Arizona), 0.08% degradation; Rocky Mountain
National Park (Colorado), 0.08% degradation; and Three Sisters Wilderness (Oregon), 0.06%
degradation. Table 7-10 contains the full visibility results from 2030 for the 138 analyzed
areas.
n The level of visibility impairment in an area is based on the light-extinction coefficient and a unitless visibility
index, called a "deciview", which is used in the valuation of visibility. The deciview metric provides a scale for
perceived visual changes over the entire range of conditions, from clear to hazy. Under many scenic conditions,
the average person can generally perceive a change of one deciview. The higher the deciview value, the worse
the visibility. Thus, an improvement in visibility is a decrease in deciview value.
7-80
-------
Environmental and Health Impacts
Table 7-10 Visibility Levels in Deciviews for Individual U.S. Class I Areas on the 20% Worst Days for
Several Scenarios
CLASS 1
AREA
(20% WORST
DAYS)
Sipsey
Wilderness
Caney Creek
Wilderness
Upper Buffalo
Wilderness
Chiricahua NM
Chiricahua
Wilderness
Galiuro
Wilderness
Grand Canyon
NP
Mazatzal
Wilderness
Mount Baldy
Wilderness
Petrified Forest
NP
Pine Mountain
Wilderness
Saguaro NM
Sierra Ancha
Wilderness
Superstition
Wilderness
Sycamore
Canyon
Wilderness
Agua Tibia
Wilderness
Ansel Adams
Wilderness
(Minarets)
Caribou
Wilderness
Cucamonga
Wilderness
Desolation
Wilderness
Emigrant
Wilderness
Hoover
Wilderness
STATE
AL
AR
AR
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
CA
CA
CA
CA
CA
CA
CA
2005
BASELINE
VISIBILITY
29.62
26.78
27.09
13.33
13.33
13.33
11.85
13.80
11.27
13.73
13.80
14.53
14.37
14.01
15.34
23.09
14.90
14.19
19.35
12.52
17.37
11.92
2030
BASE
23.41
22.52
23.06
13.28
13.27
13.20
11.58
13.10
11.10
13.31
13.12
14.04
13.82
13.46
15.04
24.56
14.78
13.98
18.23
12.54
17.21
11.85
2030
CONT-
ROL
23.41
22.51
23.05
13.28
13.27
13.20
11.58
13.10
11.10
13.32
13.12
14.04
13.82
13.46
15.04
24.55
14.77
13.97
18.22
12.54
17.20
11.85
NATURAL
BACKGROUND
11.39
11.33
11.28
6.92
6.91
6.88
6.95
6.91
6.95
6.97
6.92
6.84
6.92
6.88
6.96
7.17
7.12
7.29
7.17
7.13
7.14
7.12
7-81
-------
Regulatory Impact Analysis
John Muir
Wilderness
Joshua Tree
NM
Kaiser
Wilderness
Kings Canyon
NP
Lassen
Volcanic NP
Lava Beds NM
Mokelumne
Wilderness
Pinnacles NM
Point Reyes NS
Redwood NP
San Gabriel
Wilderness
San Gorgonio
Wilderness
San Jacinto
Wilderness
San Rafael
Wilderness
Sequoia NP
South Warner
Wilderness
Thousand
Lakes
Wilderness
Ventana
Wilderness
Yosemite NP
Black Canyon
of the Gunnison
NM
Eagles Nest
Wilderness
Flat Tops
Wilderness
Great Sand
Dunes NM
La Garita
Wilderness
Maroon Bells-
Snowmass
Wilderness
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CO
CO
CO
CO
CO
CO
14.90
19.40
14.90
23.41
14.19
14.77
12.52
18.22
22.89
18.66
19.35
21.80
21.80
19.04
23.41
14.77
14.19
18.22
17.37
10.18
9.38
9.38
12.49
10.18
9.38
14.81
18.68
14.71
22.81
14.00
14.31
12.50
18.10
22.98
19.22
18.06
20.23
20.12
18.94
22.64
14.58
14.00
18.58
17.24
9.82
9.19
9.27
12.29
10.00
9.23
14.80
18.68
14.71
22.80
14.00
14.31
12.49
18.09
22.98
19.22
18.05
20.21
20.11
18.93
22.64
14.58
14.00
18.58
17.24
9.82
9.19
9.27
12.28
10.00
9.23
7.14
7.08
7.13
7.13
7.31
7.49
7.14
7.34
7.39
7.81
7.17
7.10
7.12
7.28
7.13
7.32
7.32
7.32
7.14
7.06
7.08
7.07
7.10
7.06
7.07
7-82
-------
Environmental and Health Impacts
Mesa Verde NP
Mount Zirkel
Wilderness
Rawah
Wilderness
Rocky Mountain
NP
Weminuche
Wilderness
West Elk
Wilderness
Everglades NP
Okefenokee
Wolf Island
Craters of the
Moon NM
Sawtooth
Wilderness
Mammoth Cave
NP
Acadia NP
Moosehorn
Roosevelt
Campobello
International
Park
Isle Royale NP
Seney
Voyageurs NP
Hercules-
Glades
Wilderness
Anaconda-
Pintler
Wilderness
Bob Marshall
Wilderness
Cabinet
Mountains
Wilderness
Gates of the
Mountains
Wilderness
Glacier NP
CO
CO
CO
CO
CO
CO
FL
GA
GA
ID
ID
KY
ME
ME
ME
Ml
Ml
MN
MO
MT
MT
MT
MT
MT
12.78
10.19
10.19
13.54
10.18
9.38
22.48
27.24
27.24
14.19
14.33
31.76
23.19
21.94
21.94
21.33
24.71
19.82
27.15
13.91
14.54
14.15
11.67
19.13
12.44
10.08
9.99
13.33
9.99
9.20
21.34
23.44
23.44
13.56
14.24
25.48
22.20
21.03
21.03
19.42
22.45
17.79
23.60
13.72
14.32
13.81
11.47
18.55
12.44
10.08
9.99
13.34
9.99
9.20
21.34
23.44
23.44
13.56
14.24
25.48
22.20
21.03
21.03
19.42
22.45
17.79
23.60
13.72
14.32
13.81
11.47
18.55
7.09
7.08
7.08
7.05
7.06
7.07
11.15
11.45
11.42
7.13
7.15
11.53
11.45
11.36
11.36
11.22
11.37
11.09
11.27
7.28
7.36
7.43
7.22
7.56
7-83
-------
Regulatory Impact Analysis
Medicine Lake
Mission
Mountains
Wilderness
Scapegoat
Wilderness
Selway-
Bitterroot
Wilderness
UL Bend
Linville Gorge
Wilderness
Shining Rock
Wilderness
Lostwood
Theodore
Roosevelt NP
Great Gulf
Wilderness
Presidential
Range-Dry
River
Wilderness
Brigantine
Bandelier NM
Bosque del
Apache
Gila Wilderness
Pecos
Wilderness
Salt Creek
San Pedro
Parks
Wilderness
Wheeler Peak
Wilderness
White Mountain
Wilderness
Jarbidge
Wilderness
Wichita
Mountains
Crater Lake NP
Diamond Peak
Wilderness
MT
MT
MT
MT
MT
NC
NC
ND
ND
NH
NH
NJ
NM
NM
NM
NM
NM
NM
NM
NM
NV
OK
OR
OR
17.78
14.54
14.54
13.91
14.92
29.40
28.72
19.50
17.69
22.13
22.13
29.28
11.87
13.89
13.32
10.10
18.20
10.39
10.10
13.52
12.13
23.79
14.04
14.04
16.81
14.25
14.30
13.79
14.63
23.36
23.04
17.95
16.29
20.19
20.19
25.88
11.29
13.18
13.03
9.82
17.21
10.06
9.70
12.94
12.09
20.50
13.76
13.71
16.81
14.25
14.29
13.79
14.63
23.36
23.04
17.95
16.29
20.18
20.18
25.87
11.28
13.16
13.03
9.82
17.20
10.06
9.70
12.94
12.09
20.49
13.76
13.71
7.30
7.39
7.29
7.32
7.18
11.43
11.45
7.33
7.31
11.31
11.33
11.28
7.02
6.97
6.95
7.04
6.99
7.03
7.07
6.98
7.10
11.07
7.71
111
7-84
-------
Environmental and Health Impacts
Eagle Cap
Wilderness
Gearhart
Mountain
Wilderness
Hells Canyon
Wilderness
Kalmiopsis
Wilderness
Mount Hood
Wilderness
Mount Jefferson
Wilderness
Mount
Washington
Wilderness
Mountain Lakes
Wilderness
Strawberry
Mountain
Wilderness
Three Sisters
Wilderness
Cape Romain
Badlands NP
Wind Cave NP
Great Smoky
Mountains NP
Joyce-Kilmer-
Slickrock
Wilderness
Big Bend NP
Carlsbad
Caverns NP
Guadalupe
Mountains NP
Arches NP
Bryce Canyon
NP
Canyonlands
NP
Capitol Reef NP
James River
Face
Wilderness
Shenandoah
NP
Lye Brook
Wilderness
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
SC
SD
SD
TN
TN
TX
TX
TX
UT
UT
UT
UT
VA
VA
VT
18.25
14.04
18.73
16.31
14.79
15.93
15.93
14.04
18.25
15.93
27.14
16.73
15.96
30.43
30.43
17.39
16.98
16.98
11.04
11.73
11.04
10.63
29.32
29.66
24.17
17.64
13.88
17.90
16.38
14.49
15.75
15.72
13.75
17.65
15.72
24.09
15.52
14.93
24.30
24.30
16.43
15.89
15.89
10.82
11.52
10.88
10.74
23.18
23.73
20.72
17.63
13.88
17.91
16.39
14.49
15.75
15.72
13.74
17.66
15.73
24.09
15.51
14.93
24.30
24.30
16.42
15.88
15.88
10.81
11.52
10.88
10.74
23.17
23.72
20.72
7.34
7.46
7.32
7.71
111
7.81
7.89
7.57
7.49
7.87
11.36
7.30
7.24
11.44
11.45
6.93
7.02
7.03
6.99
6.99
7.01
7.03
11.24
11.25
11.25
7-85
-------
Regulatory Impact Analysis
Alpine Lake
Wilderness
Glacier Peak
Wilderness
Goat Rocks
Wilderness
Mount Adams
Wilderness
Mount Rainier
NP
North Cascades
NP
Olympic NP
Pasayten
Wilderness
Dolly Sods
Wilderness
Otter Creek
Wilderness
Bridger
Wilderness
Fitz patrick
Wilderness
Grand Teton
NP
North Absaroka
Wilderness
Red Rock
Lakes
Teton
Wilderness
Washakie
Wilderness
Yellowstone NP
WA
WA
WA
WA
WA
WA
WA
WA
WV
WV
WY
WY
WY
WY
WY
WY
WY
WY
17.35
13.78
12.88
12.88
17.56
13.78
16.14
15.39
29.73
29.73
10.93
10.93
10.94
11.12
10.94
10.94
11.12
10.94
17.29
14.06
12.32
12.33
17.23
14.20
16.35
14.99
23.14
23.14
10.80
10.80
10.61
10.98
10.68
10.70
10.98
10.66
17.28
14.05
12.32
12.33
17.22
14.19
16.35
14.99
23.14
23.14
10.80
10.80
10.61
10.98
10.68
10.70
10.98
10.66
7.86
7.80
7.82
7.78
7.90
7.78
7.88
111
11.32
11.33
7.08
7.09
7.09
7.09
7.14
7.09
7.09
7.12
7.3 Quantified and Monetized Non-GHG Health and Environmental
Impacts
This section presents EPA's analysis of the non-GHG health and environmental
impacts that can be expected to occur as a result of the light-duty vehicle GHG rule. GHG
emissions are predominantly the byproduct of fossil fuel combustion processes that also
produce criteria and hazardous air pollutants. The vehicles that are subject to the standards
are also significant sources of mobile source air pollution such as direct PM, NOx, VOCs and
air toxics. The standards will affect exhaust emissions of these pollutants from vehicles.
They will also affect emissions from upstream sources related to changes in fuel consumption.
7-86
-------
Environmental and Health Impacts
Changes in ambient ozone, PM2.5, and air toxics that will result from the standards are
expected to affect human health in the form of premature deaths and other serious human
health effects, as well as other important public health and welfare effects.
It is important to quantify the health and environmental impacts associated with the
final rule because a failure to adequately consider these ancillary co-pollutant impacts could
lead to an incorrect assessment of their net costs and benefits. Moreover, co-pollutant impacts
tend to accrue in the near term, while any effects from reduced climate change mostly accrue
over a time frame of several decades or longer.
This section is split into two sub-sections: the first presents the PM- and ozone-related
health and environmental impacts associated with final rule in calendar year (CY) 2030; the
second presents the PM-related benefits-per-ton values used to monetize the PM-related co-
benefits associated with the model year (MY) analysis of the final rule/1
7.3.1 Quantified and Monetized Non-GHG Human Health Benefits of the 2030
Calendar Year (CY) Analysis
This analysis reflects the impact of the final light-duty GHG rule in 2030 compared to
a future-year reference scenario without the rule in place. Overall, we estimate that the final
rule will lead to a net decrease in PMi.s-related health impacts (see Chapter 7.2.2.2 for more
information about the air quality modeling results). While the PM-related air quality impacts
are relatively small, the decrease in population-weighted national average PMi.5 exposure
results in a net decrease in adverse PM-related human health impacts (the decrease in national
population weighted annual average PM2.5 is 0.0036
The air quality modeling also projects very small increases in ozone concentrations in
many areas (see Chapter 7.2.2.1), but these are driven by the ethanol production volumes
mandated by the recently finalized RFS2 rule and are not due to the standards finalized in this
rule (see Chapters 5.3.2 and 5.3.3.5 for more information). While the ozone-related impacts
are very small, the overall increase in population-weighted national average ozone exposure
results in a small increase in ozone-related health impacts (population weighted maximum 8-
hour average ozone increases by 0.0104 ppb).
JJ EPA typically analyzes rule impacts (emissions, air quality, costs and benefits) in the year in which they occur;
for this analysis, we selected 2030 as a representative future year. We refer to this analysis as the "Calendar
Year" (CY) analysis. EPA also conducted a separate analysis of the impacts over the model year lifetimes of the
2012 through 2016 model year vehicles. We refer to this analysis as the "Model Year" (MY) analysis. In
contrast to the CY analysis, the MY lifetime analysis shows the lifetime impacts of the program on each of these
MY fleets over the course of its lifetime.
7-87
-------
Regulatory Impact Analysis
We base our analysis of the final rule's impact on human health in 2030 on peer-
reviewed studies of air quality and human health effects.274'275 Our benefits methods are also
consistent with recent rulemaking analyses such as the proposed Portland Cement National
Emissions Standards for Hazardous Air Pollutants (NESHAP) RIA,276 the final NO2 NAAQS,
277 and the final Category 3 Marine Engine rule.278 To model the ozone and PM air quality
impacts of the final rule, we used the Community Multiscale Air Quality (CMAQ) model (see
Section 7.2.1). The modeled ambient air quality data serves as an input to the Environmental
Benefits Mapping and Analysis Program (BenMAP).KK BenMAP is a computer program
developed by the U.S. EPA that integrates a number of the modeling elements used in
previous analyses (e.g., interpolation functions, population projections, health impact
functions, valuation functions, analysis and pooling methods) to translate modeled air
concentration estimates into health effects incidence estimates and monetized benefits
estimates.
The range of total monetized ozone- and PM-related health impacts is presented in
Table 7-11. We present total benefits based on the PM- and ozone-related premature
mortality function used. The benefits ranges therefore reflect the addition of each estimate of
ozone-related premature mortality (each with its own row in Table 7-11) to estimates of PM-
related premature mortality. These estimates represent EPA's preferred approach to
characterizing a best estimate of benefits. As is the nature of Regulatory Impact Analyses
(RIAs), the assumptions and methods used to estimate air quality benefits evolve to reflect the
Agency's most current interpretation of the scientific and economic literature.
Table 7-11 Estimated 2030 Monetized PM-and Ozone-Related Health Benefits3
2030 Total Ozone and PM Benefits - PM Mortality Derived from American Cancer Society Analysis and
Six-Cities Analysis2
Premature Ozone
Mortality Function
Multi-city analyses
Meta-analyses
Reference
Bell et al., 2004
Huang et al., 2005
Schwartz, 2005
Bell et al., 2005
Total Benefits
(Millions, 2007$, 3%
Discount Rate)b'c'd
Total: $510 -$1,300
PM: $550 - $1,300
Ozone: -$40
Total: $490 -$1,300
PM: $550 - $1,300
Ozone: -$64
Total: $490 -$1,300
PM: $550 - $1,300
Ozone: -$60
Total: $430 -$1,200
PM: $550 - $1,300
Ozone: -$120
Total Benefits
(Millions, 2007$, 7%
Discount Rate) b'c'd
Total: $460 -$1,200
PM: $500 - $1,200
Ozone: -$40
Total: $440 -$1,200
PM: $500 - $1,200
Ozone: -$64
Total: $440 -$1,200
PM: $500 - $1,200
Ozone: -$60
Total: $380 -$1,100
PM: $500 - $1,200
Ozone: -$120
KK Information on BenMAP, including downloads of the software, can be found at http://www.epa.gov/ttn/ecas/
benmodels. html.
7-88
-------
Environmental and Health Impacts
Ito et al., 2005
Levy etal., 2005
Total: $380 -$1,200
PM: $550 - $1,300
Ozone: -$170
Total: $380 -$1,200
PM: $550 - $1,300
Ozone: -$170
Total: $330 -$1,000
PM: $500 - $1,200
Ozone: -$170
Total: $330 -$1,000
PM: $500 - $1,200
Ozone: -$170
Notes:
aTotal includes premature mortality-related and morbidity-related ozone and PM2.5benefits. Range was
developed by adding the estimate from the ozone premature mortality function to the estimate of PM2.5-
related premature mortality derived from either the ACS study (Pope et al., 2002) or the Six-Cities
study (Laden et al., 2006).
b Note that total benefits presented here do not include a number of unquantified benefits categories. A
detailed listing of unquantified health and welfare effects is provided in Table 7-12.
c Results reflect the use of both a 3 and 7 percent discount rate, as recommended by EPA's Guidelines
for Preparing Economic Analyses and OMB Circular A-4. Results are rounded to two significant digits
for ease of presentation and computation.
d Negatives indicate a disbenefit, or an increase in health effect incidence.
The benefits in Table 7-11 include all of the human health impacts we are able to
quantify and monetize at this time. However, the full complement of human health and
welfare effects associated with PM and ozone remain unquantified because of current
limitations in methods or available data. We have not quantified a number of known or
suspected health effects linked with ozone and PM for which appropriate health impact
functions are not available or which do not provide easily interpretable outcomes (e.g.,
changes in heart rate variability). Additionally, we are unable to quantify a number of known
welfare effects, including reduced acid and particulate deposition damage to cultural
monuments and other materials, and environmental benefits due to reductions of impacts of
eutrophication in coastal areas. These are listed in Table 7-12. As a result, the health benefits
quantified in this section are likely underestimates of the total benefits attributable to the final
rule.
Table 7-12 Unquantified and Non-Monetized Potential Effects
Pollutant/Effects
Ozone Health
Ozone Welfare
PM Health0
Effects Not Included in Analysis - Changes in:
Chronic respiratory damage"
Premature aging of the lungsb
Non-asthma respiratory emergency room visits
Exposure to UVb (+/-f
Yields for
-commercial forests
-some fruits and vegetables
-non-commercial crops
Damage to urban ornamental plants
Impacts on recreational demand from damaged forest aesthetics
Ecosystem functions
Exposure to UVb (+/-)e
Premature mortality - short term exposures
Low birth weight
7-89
-------
Regulatory Impact Analysis
PM Welfare
Nitrogen and Sulfate
Deposition Welfare
CO Health
HC/Toxics Health1
HC/Toxics Welfare
Pulmonary function
Chronic respiratory diseases other than chronic bronchitis
Non-asthma respiratory emergency room visits
Exposure to UVb (+/-)e
Residential and recreational visibility in non-Class I areas
Soiling and materials damage
Damage to ecosystem functions
Exposure to UVb (+/-)e
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)
Anemia (benzene)
Disruption of production of blood components (benzene)
Reduction in the number of blood platelets (benzene)
Excessive bone marrow formation (benzene)
Depression of lymphocyte counts (benzene)
Reproductive and developmental effects (1,3-butadiene)
Irritation of eyes and mucus membranes (formaldehyde)
Respiratory irritation (formaldehyde)
Asthma attacks in asthmatics (formaldehyde)
Asthma-like symptoms in non-asthmatics (formaldehyde)
Irritation of the eyes, skin, and respiratory tract (acetaldehyde)
Upper respiratory tract irritation and congestion (acrolein)
Direct toxic effects to animals
Bioaccumulation in the food chain
Damage to ecosystem function
Odor
Notes:
inflammation in the lung, acute inflammation and respiratory cell damage, and increased susceptibility
to respiratory infection are likely partially represented by our quantified endpoints.
b The public health impact of effects such as chronic respiratory damage and premature aging of the
lungs may be partially represented by quantified endpoints such as hospital admissions or premature
mortality, but a number of other related health impacts, such as doctor visits and decreased athletic
performance, remain unquantified.
c In addition to primary economic endpoints, there are a number of biological responses that have been
associated with PM health effects including morphological changes and altered host defense
mechanisms. The public health impact of these biological responses may be partly represented by our
quantified endpoints.
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
7-90
-------
Environmental and Health Impacts
analysis. However, the PM mortality results derived from the expert elicitation do take into account
premature mortality effects of short term exposures.
e May result in benefits or disbenefits.
Many of the key hydrocarbons related to this rule are also hazardous air pollutants listed in the CAA.
While there will be impacts associated with air toxic pollutant emission changes that
result from the final rule, we do not attempt to monetize those impacts. This is primarily
because currently available tools and methods to assess air toxics risk from mobile sources at
the national scale are not adequate for extrapolation to incidence estimations or benefits
assessment. The best suite of tools and methods currently available for assessment at the
national scale are those used in the National-Scale Air Toxics Assessment (NATA). The EPA
Science Advisory Board specifically commented in their review of the 1996 NATA that these
tools were not yet 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.279
While EPA has since improved the tools, there remain critical limitations for estimating
incidence and assessing benefits of reducing mobile source air toxics. EPA continues to work
to address these limitations; however, we did not have the methods and tools available for
national-scale application in time for the analysis of the final rule.LL
EPA is also unaware of specific information identifying any effects on listed
endangered species from the small fluctuations in pollutant concentrations associated with this
rule (see Section 7.2). Furthermore, our current modeling tools are not designed to trace
fluctuations in ambient concentration levels to potential impacts on particular endangered
species.
7.3.1.1 Human Health and Environmental Impacts
Table 7-13 and Table 7-14 present the annual PM2.5 and ozone health impacts in the 48
contiguous U.S. states associated with the final rule for 2030. For each endpoint presented in
Table 7-13 and Table 7-14, we provide both the point estimate and the 90% confidence
interval.
Using EPA's preferred estimates, based on the American Cancer Society (ACS) and
Six-Cities studies and no threshold assumption in the model of mortality, we estimate that the
LL 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, the workshop 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.
7-91
-------
Regulatory Impact Analysis
final rule will result in between 60 and 150 cases of avoided PM2.5-related premature deaths
annually in 2030. As a sensitivity analysis, when the range of expert opinion is used, we
estimate between 22 and 200 fewer premature mortalities in 2030.
The range of ozone impacts is based on changes in risk estimated using several
sources of ozone-related mortality effect estimates. This analysis presents six alternative
estimates for the association based upon different functions reported in the scientific
literature, derived from both the National Morbidity, Mortality, and Air Pollution Study
(NMMAPS) (Bell et al, 2004; Huang et al., 2005; Schwartz, 2005) and from a series of recent
meta-analyses (Bell et al., 2005, Ito et al., 2005, and Levy et al., 2005). This approach is not
inconsistent with recommendations provided by the NRC in their recent report (NRC, 2008)
on the estimation of ozone-related mortality risk reductions, "The committee recommends
that the greatest emphasis be placed on estimates from new systematic multicity analyses that
use national databases of air pollution and mortality, such as in the NMMAPS, without
excluding consideration of meta-analyses of previously published studies." For ozone-related
premature mortality in 2030, we estimate a range of between 4 to 18 additional premature
mortalities related to the ethanol production volumes mandated by the recently finalized RFS2
ruleMM (and reflected in the air quality modeling for this rule), but are not due to the final
standards themselves.
Following these tables, we also provide a more comprehensive presentation of the
distributions of incidence generated using the available information from empirical studies
and expert elicitation. Table 7-15 presents the distributions of the reduction in PM2.5-related
premature mortality based on the C-R distributions provided by each expert, as well as that
from the data-derived health impact functions, based on the statistical error associated with
the ACS study (Pope et al., 2002) and the Six-Cities study (Laden et al., 2006). The 90%
confidence interval for each separate estimate of PM-related mortality is also provided.
In 2030, the effect estimates of nine of the twelve experts included in the elicitation
panel fall within the empirically-derived range provided by the ACS and Six-Cities studies.
Only one expert falls below this range, while two of the experts are above this range.
Although the overall range across experts is summarized in these tables, the full uncertainty in
the estimates is reflected by the results for the full set of 12 experts. The twelve experts'
judgments as to the likely mean effect estimate are not evenly distributed across the range
illustrated by arraying the highest and lowest expert means.
MM EPA 2010, Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis. EPA-420-R-10-006.
February 2010. Docket EPA-HQ-OAR-2009-0472-11332. see also 75 FR 14670, March 26, 2010.
7-92
-------
Environmental and Health Impacts
Table 7-13 Estimated PM2.s-Related Health Impacts3
Health Effect
Premature Mortality - Derived from epidemiology literature13
Adult, age 30+, ACS Cohort Study (Pope et al., 2002)
Adult, age 25+, Six-Cities Study (Laden et al., 2006)
Infant, age <1 year (Woodruff et al., 1997)
Chronic bronchitis (adult, age 26 and over)
Non-fatal myocardial infarction (adult, age 18 and over)
Hospital admissions - respiratory (all ages)0
Hospital admissions - cardiovascular (adults, age >18)d
Emergency room visits for asthma (age 18 years and younger)
Acute bronchitis, (children, age 8-12)
Lower respiratory symptoms (children, age 7-14)
Upper respiratory symptoms (asthmatic children, age 9-18)
Asthma exacerbation (asthmatic children, age 6-18)
Work loss days
Minor restricted activity days (adults age 18-65)
2030 Annual Reduction in
Incidence
(5*% - 95ttl%ile)
60
(23 - 96)
150
(83 - 220)
0
(0-1)
42
(8-77)
100
(38 - 170)
13
(7-20)
32
(23 - 38)
42
(25 - 59)
95
(0 - 190)
1,100
(540-1,700)
850
(270-1,400)
1,000
(120-2,900)
7,600
(6,600 - 8,500)
45,000
(38,000 - 52,000)
Notes:
a Incidence is rounded to two significant digits. Estimates represent incidence within the 48 contiguous
United States.
b PM-related adult mortality based upon the American Cancer Society (ACS) Cohort Study (Pope et al.,
2002) and the Six-Cities Study (Laden et al., 2006). Note that these are two alternative estimates of
adult mortality and should not be summed. PM-related infant mortality based upon a study by
Woodruff, Grillo, and Schoendorf, (1997).280
c Respiratory hospital admissions for PM include admissions for chronic obstructive pulmonary disease
(COPD), pneumonia and asthma.
d Cardiovascular hospital admissions for PM include total cardiovascular and subcategories for ischemic
heart disease, dysrhythmias, and heart failure.
7-93
-------
Regulatory Impact Analysis
Table 7-14 Estimated Ozone-Related Health Impacts3
Health Effect
Premature Mortality, All ages"
Multi-City Analyses
Bell et al. (2004) - Non-accidental
Huang et al. (2005) - Cardiopulmonary
Schwartz (2005) - Non-accidental
Meta-analyses:
Bell et al. (2005) - All cause
Ito et al. (2005) - Non-accidental
Levy et al. (2005) - All cause
Hospital admissions- respiratory causes (adult, 65 and
older)c
Hospital admissions -respiratory causes (children, under
2)
Emergency room visit for asthma (all ages)
Minor restricted activity days (adults, age 18-65)
School absence days
2030 Annual Reduction in
Incidence
(5th% - 95th%ile)
-4
(-8-0)
-7
(-14-1)
-6
(-13-1)
-13
(-24 - -2)
-18
(-30 - -6)
-18
(-28 - -9)
-38
(-86 - -6)
-6
(-13-1)
-16
(-51-8)
-18,000
(-40,000 - 3,700)
-7,700
(-16,000-1,200)
Notes:
a Negatives indicate a disbenefit, or an increase in health effect incidence. Incidence is rounded to two
significant digits. Estimates represent incidence within the 48 contiguous U.S.
b Estimates of ozone-related premature mortality are based upon incidence estimates derived from several
alternative studies: Bell et al. (2004); Huang et al. (2005); Schwartz (2005); Bell et al. (2005); Ito et al.
(2005); Levy et al. (2005). The estimates of ozone-related premature mortality should therefore not be
summed.
c Respiratory hospital admissions for ozone include admissions for all respiratory causes and
subcategories for COPD and pneumonia.
7-94
-------
Environmental and Health Impacts
Table 7-15 Results of Application of Expert Elicitation: Annual Reductions in Premature Mortality in
2030 Associated with the Final Rule
Source of Mortality
Estimate
Pope et al. (2002)
Laden et al. (2006)
Expert A
Expert B
Expert C
Expert D
Expert E
Expert F
Expert G
Expert H
Expert I
Expert J
Expert K
Expert L
2030 Primary Option
5th Percentile
23
83
30
17
22
18
100
78
0
0
19
29
0
14
Mean
60
150
160
120
120
86
200
110
72
91
120
98
22
87
95th Percentile
96
220
300
270
270
140
310
160
130
210
220
220
100
170
7.3.1.2 Monetized Estimates of Impacts of Changes in Non-GHG Pollutants
Table 7-16 presents the estimated monetary value of changes in the incidence of ozone
and PMi.s-related health effects. Total aggregate monetized benefits are presented in Table
7-17. All monetized estimates are presented in 2007$. Where appropriate, estimates account
for growth in real gross domestic product (GDP) per capita between 2000 and 2030.NN The
monetized value of PM2.5-related mortality also accounts for a twenty-year segmented
Our analysis accounts for expected growth in real income over time. Economic theory argues that WTP for
most goods (such as environmental protection) will increase if real incomes increase. Benefits are therefore
adjusted by multiplying the unadjusted benefits by the appropriate adjustment factor to account for income
growth over time. For growth between 2000 and 2030, this factor is 1.23 for long-term mortality, 1.27 for
chronic health impacts, and 1.08 for minor health impacts. For a complete discussion of how these adjustment
factors were derived, we refer the reader to the PM NAAQS regulatory impact analysis.9 Note that similar
adjustments do not exist for cost-of-illness-based unit values. For these, we apply the same unit value regardless
of the future year of analysis.
7-95
-------
Regulatory Impact Analysis
cessation lag.00 To discount the value of premature mortality that occurs at different points in
the future, we apply both a 3% and 7% discount rate. We also use both a 3% and 7% discount
rate to value PM-related nonfatal heart attacks (myocardial infarctions).pp As the results
indicate, total benefits are driven primarily by the reduction in PMi.s-related premature
fatalities each year.
In addition to omitted benefits categories such as air toxics and various welfare
effects, not all known PM2.5- and ozone-related health and welfare effects could be quantified
or monetized. The estimate of total monetized health benefits of the final rule is thus equal to
the subset of monetized PM2.5- and ozone-related health impacts we are able to quantify plus
the sum of the nonmonetized health and welfare benefits. Our estimate of total monetized
benefits in 2030 for the final rule, using the ACS and Six-Cities PM mortality studies and the
range of ozone mortality assumptions, is between $380 and $1,300 million, assuming a 3
percent discount rate, or between $330 and $1,200 million, assuming a 7 percent discount
rate. As the results indicate, total benefits are driven primarily by the reduction in PM2.5-
related premature fatalities each year.
The next largest benefit is for reductions in chronic illness (chronic bronchitis and
nonfatal heart attacks), although this value is more than an order of magnitude lower than for
premature mortality. Hospital admissions for respiratory and cardiovascular causes, minor
restricted activity days, and work loss days account for the majority of the remaining benefits.
The remaining categories each account for a small percentage of total benefit; however, they
represent a large number of avoided incidences affecting many individuals. A comparison of
the incidence table to the monetary benefits table reveals that there is not always a close
correspondence between the number of incidences avoided for a given endpoint and the
monetary value associated with that endpoint. For example, there are over 100 times more
work loss days than PM-related premature mortalities (based on the ACS study), yet work
loss days account for only a very small fraction of total monetized benefits. This reflects the
fact that many of the less severe health effects, while more common, are valued at a lower
level than the more severe health effects. Also, some effects, such as hospital admissions, are
00 Based in part on prior SAB advice, EPA has typically assumed that there is a time lag between changes in
pollution exposures and the total realization of changes in health effects. Within the context of benefits analyses,
this term is often referred to as "cessation lag". The existence of such a lag is important for the valuation of
premature mortality incidence because economic theory suggests that benefits occurring in the future should be
discounted. In this analysis, we apply a twenty-year distributed lag to PM mortality reductions. This method is
consistent with the most recent recommendation by the EPA's Science Advisory Board. Refer to: EPA -
Science Advisory Board, 2004. Advisory Council on Clean Air Compliance Analysis Response to Agency
Request on Cessation Lag. Letter from the Health Effects Subcommittee to the U.S. Environmental Protection
Agency Administrator, December.
pp Nonfatal myocardial infarctions (MI) are valued using age-specific cost-of-illness values that reflect lost
earnings and direct medical costs over a 5-year period following a nonfatal MI.
7-96
-------
Environmental and Health Impacts
valued using a proxy measure of willingness-to-pay (e.g., cost-of-illness). As such, the true
value of these effects may be higher than that reported here.
Table 7-16 Estimated Monetary Value of Changes in Incidence of Health and Welfare Effects (in millions
of2007$)a'b
PM2 s-Related Health Effect
Premature Mortality -
Derived from Epidemiology
Studies0' '
Adult, age 30+ - ACS study
(Pope et al., 2002)
3% discount rate
7% discount rate
Adult, age 25+ - Six-Cities study
(Laden et al., 2006)
3% discount rate
7% discount rate
Infant Mortality, <1 year -
(Woodruff etal. 1997)
Chronic bronchitis (adults, 26 and over)
Non-fatal acute myocardial infarctions
3% discount rate
7% discount rate
Hospital admissions for respiratory causes
Hospital admissions for cardiovascular causes
Emergency room visits for asthma
Acute bronchitis (children, age 8-12)
Lower respiratory symptoms (children, 7-14)
Upper respiratory symptoms (asthma, 9-1 1)
Asthma exacerbations
Work loss days
2030
(5th and 95th %ile)
$510
($70 -$1,300)
$460
($63 - $1,200)
$1,300
($190 -$3,300)
$1,200
($180 -$3,000)
$1.8
($0 - $7.0)
$22
($1.9 - $77)
$14
($3.9 - $35)
$14
($3.6 -$35)
$0.20
($0.01 - $0.29)
$0.91
($0.58 -$1.3)
$0.016
($0.009 - $0.024)
$0.007
($0- $0.018)
$0.022
($0.009 - $0.043)
$0.027
($0.008 - $0.061)
$0.058
($0.006 -$0.17)
$1.2
($1.0 -$1.3)
7-97
-------
Regulatory Impact Analysis
Minor restricted-activity days (MRADs)
Ozone-related Health Effect
Premature Mortality, All ages -
Derived from Multi-city analyses
Premature Mortality, All ages -
Derived from Meta-analyses
Bell etal., 2004
Huang etal., 2005
Schwartz, 2005
Bell etal., 2005
Ito etal., 2005
Levy et al., 2005
Hospital admissions- respiratory causes (adult, 65 and older)
Hospital admissions- respiratory causes (children, under 2)
Emergency room visit for asthma (all ages)
Minor restricted activity days (adults, age 18-65)
School absence days
$2.9
($1.7 - $4.2)
(5th and 95th %ile)
-$38
(-$110 -$4.2)
-$62
(-$180 -$4.7)
-$58
(-$170 -$8.8)
-$120
(-$330 - -$7.9)
-$170
(-$430 --$19)
-$170
(-$410 - -$21)
-$0.92
(-$2.1 -$0.27)
-$.21
(-$.45 -$0.031)
-$0.006
(-$0.018 - $0.003)
-$1.2
(-$2.7 - $0.25)
-$0.71
(-$1.4 -$0.11)
Notes:
a Negatives indicate a disbenefit, or an increase in health effect incidence. Monetary benefits are
rounded to two significant digits for ease of presentation and computation. PM and ozone benefits are
nationwide.
b Monetary benefits adjusted to account for growth in real GDP per capita between 1990 and the
analysis year (2030).
c Valuation assumes discounting over the SAB recommended 20 year segmented lag structure. Results
reflect the use of 3 percent and 7 percent discount rates consistent with EPA and OMB guidelines for
preparing economic analyses.
7-98
-------
Environmental and Health Impacts
Table 7-17 Total Monetized Ozone and PM-related Benefits Associated with the Final Rule in 2030
Total Ozone and PM Benefits (billions, 2007$) -
PM Mortality Derived from the ACS and Six-Cities Studies
3% Discount Rate
Ozone
Mortality
Function
Multi-city
Meta-analysis
PM
Reference
Bell et al.,
2004
Huang et al.,
2005
Schwartz,
2005
Bell et al.,
2005
Ito et al., 2005
Levy etal.,
2005
Total
Mean Total
Benefits
$510 -$1,300
$490 - $1,300
$490 - $1,300
$430 - $1,200
$380 - $1,200
$380 - $1,200
7% Discount Rate
Ozone
Mortality
Function
Multi-city
Meta-analysis
Ozone and PM Benefits (billions,
Reference
Bell etal.,
2004
Huang etal.,
2005
Schwartz,
2005
Bell etal.,
2005
Ito etal., 2005
Levy et al.,
2005
2007$) -
Mean Total
Benefits
$460 - $1,200
$440 - $1,200
$440 - $1,200
$380 -$1,100
$330 - $1,000
$330 - $1,000
Mortality Derived from Expert Elicitation (Lowest and Highest Estimate)
3% Discount Rate
Ozone
Mortality
Function
Multi-city
Meta-analysis
Reference
Bell et al.,
2004
Huang et al.,
2005
Schwartz,
2005
Bell et al.,
2005
Ito et al., 2005
Levy etal.,
2005
Mean Total
Benefits
$190 - $1,700
$160 - $1,700
$170 - $1,700
$100 - $1,600
$56 - $1,600
$55 -$1,600
7% Discount Rate
Ozone
Mortality
Function
Multi-city
Meta-analysis
Reference
Bell etal.,
2004
Huang etal.,
2005
Schwartz,
2005
Bell etal.,
2005
Ito etal., 2005
Levy et al.,
2005
Mean Total
Benefits
$170 - $1,600
$140 - $1,500
$150 -$1,500
$86 - $1,500
$37 - $1,400
$36 - $1,400
7.3.1.3 Methodology
7.3.1.3.1 Human Health Impact Functions
Health impact functions measure the change in a health endpoint of interest, such as
hospital admissions, for a given change in ambient ozone or PM concentration. Health impact
functions are derived from primary epidemiology studies, meta-analyses of multiple
epidemiology studies, or expert elicitations. A standard health impact function has four
7-99
-------
Regulatory Impact Analysis
components: (1) an effect estimate from a particular study; (2) a baseline incidence rate for
the health effect (obtained from either the epidemiology study or a source of public health
statistics such as the Centers for Disease Control); (3) the size of the potentially affected
population; and (4) the estimated change in the relevant ozone or PM summary measures.
A typical health impact function might look like:
where yo is the baseline incidence (the product of the baseline incidence rate times the
potentially affected population), p is the effect estimate, and Ax is the estimated change in the
summary pollutant measure. There are other functional forms, but the basic elements remain
the same. The following subsections describe the sources for each of the first three elements:
size of the potentially affected populations; PM2.5 and ozone effect estimates; and baseline
incidence rates. We also describe the treatment of potential thresholds in PM-related health
impact functions. Section 7.2 describes the ozone and PM air quality inputs to the health
impact functions.
7.3.1.3.1.1 Potentially Affected Populations
The starting point for estimating the size of potentially affected populations is the
2000 U.S. Census block level dataset.281 Benefits Modeling and Analysis Program
(BenMAP) incorporates 250 age/gender/race categories to match specific populations
potentially affected by ozone and other air pollutants. The software constructs specific
populations matching the populations in each epidemiological study by accessing the
appropriate age-specific populations from the overall population database. BenMAP projects
populations to 2030 using growth factors based on economic projections.282
7.3.1.3.1.2 Effect Estimate Sources
The most significant quantifiable benefits of reducing ambient concentrations of ozone
and PM are attributable to reductions in human health risks. EPA's Ozone and PM Criteria
Documents283'284 and the World Health Organization's 2003 and 2004285'286 reports outline
numerous human health effects known or suspected to be linked to exposure to ambient ozone
and PM. EPA recently evaluated the ozone and PM literature for use in the benefits analysis
for the final 2008 Ozone NAAQS and final 2006 PM NAAQS analyses. We use the same
literature in this analysis; for more information on the studies that underlie the health impacts
quantified in this RIA, please refer to those documents.
It is important to note that we are unable to separately quantify all of the possible PM
and ozone health effects that have been reported in the literature for three reasons: (1) the
possibility of double counting (such as hospital admissions for specific respiratory diseases
7-100
-------
Environmental and Health Impacts
versus hospital admissions for all or a sub-set of respiratory diseases); (2) uncertainties in
applying effect relationships that are based on clinical studies to the potentially affected
population; or (3) the lack of an established concentration-response (CR) relationship. Table
7-18 lists the health endpoints included in this analysis.
Table 7-18 Health Impact Functions Used in BenMAP to Estimate Impacts
Reductions
and Ozone
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)287 - Non-
accidental
Huang et al (2005)288 - Cardiopulmonary
Schwartz (2005)289 - Non-accidental
Meta-analyses:
Bell et al (2005)290 - All cause
Ito et al (2005)291 - Non-accidental
Levy et al (2005)292 - All cause
Pope et al. (2002)293
Laden et al. (2006)294
Expert Elicitation (lEc, 2006)295
Woodruff etal. (1997)296
All ages
>29 years
>25 years
>24 years
Infant (<1 year)
Chronic Illness
Chronic bronchitis
Nonfatal heart attacks
PM2.5
PM2.5
Abbey etal. (1995)297
Peters etal. (200 1)298
>26 years
Adults (>18 years)
Hospital Admissions
Respiratory
03
PM2.5
Pooled estimate:
Schwartz (1995) - ICD 460-519 (all resp)299
Schwartz (1994a; 1994b) - ICD 480-486
(pneumonia)300'301
Moolgavkar et al. (1997) - ICD 480-487
(pneumonia)302
Schwartz (1994b) - ICD 491-492, 494-496
(COPD)
Moolgavkar et al. (1997) - ICD 490-496
(COPD)
Burnett etal. (2001)303
Pooled estimate:
Moolgavkar (2003)— ICD 490-496 (COPD)304
Ito (2003)— ICD 490-496 (COPD)305
>64 years
<2 years
>64 years
7-101
-------
Regulatory Impact Analysis
Cardiovascular
Asthma-related ER
visits
Asthma-related ER
visits (con't)
PM2.5
PM2.5
PM2.5
PM2.5
PM2.5
03
PM2.5
Moolgavkar (2000)— ICD 490-496 (COPD)306
Ito (2003)— ICD 480-486 (pneumonia)
Sheppard (2003)— ICD 493 (asthma)307
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)
Pooled estimate:
Jaffe et al (2003)308
Peel et al (2005)309
Wilson et al (2005)310
Norrisetal. (1999)311
20-64 years
>64 years
<65 years
>64 years
20-64 years
5-34 years
All ages
All ages
0-18 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)312
Popeetal. (1991)313
Schwartz andNeas (2000)314
Pooled estimate:
Ostro et al. (2001)315 (cough, wheeze and
shortness of breath)
Vedal et al. (1998)316 (cough)
Ostro (1987)317
Pooled estimate:
Gilliland et al. (2001)318
Chen et al. (2000)319
Ostro and Rothschild (1989)320
Ostro and Rothschild (1989)
8-12 years
Asthmatics, 9-11
years
7-14 years
6-18 years2
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 Gilliland et al. (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.
7-102
-------
Environmental and Health Impacts
In selecting epidemiological studies as sources of effect estimates, we applied several
criteria to develop a set of studies that is likely to provide the best estimates of impacts in the
U.S. To account for the potential impacts of different health care systems or underlying
health status of populations, we give preference to U.S. studies over non-U.S. studies. In
addition, due to the potential for confounding by co-pollutants, we give preference to effect
estimates from models including both ozone and PM over effect estimates from single-
pollutant models.321'322
7.3.1.3.1.3 Baseline Incidence Rates
Epidemiological studies of the association between pollution levels and adverse health
effects generally provide a direct estimate of the relationship of air quality changes to the
relative risk of a health effect, rather than estimating the absolute number of avoided cases.
For example, a typical result might be that a 100 ppb decrease in daily ozone levels might, in
turn, decrease hospital admissions by 3 percent. The baseline incidence of the health effect is
necessary to convert this relative change into a number of cases. A baseline incidence rate is
the estimate of the number of cases of the health effect per year in the assessment location, as
it corresponds to baseline pollutant levels in that location. To derive the total baseline
incidence per year, this rate must be multiplied by the corresponding population number. For
example, if the baseline incidence rate is the number of cases per year per 100,000 people,
that number must be multiplied by the number of 100,000s in the population.
Table 7-19 summarizes the sources of baseline incidence rates and provides average
incidence rates for the endpoints included in the analysis. For both baseline incidence and
prevalence data, we used age-specific rates where available. We applied concentration-
response functions to individual age groups and then summed over the relevant age range to
provide an estimate of total population benefits. In most cases, we used a single national
incidence rate, due to a lack of more spatially disaggregated data. Whenever possible, the
national rates used are national averages, because these data are most applicable to a national
assessment of benefits. For some studies, however, the only available incidence information
comes from the studies themselves; in these cases, incidence in the study population is
assumed to represent typical incidence at the national level. Regional incidence rates are
available for hospital admissions, and county-level data are available for premature mortality.
We have projected mortality rates such that future mortality rates are consistent with our
projections of population growth.323
7-103
-------
Regulatory Impact Analysis
Table 7-19 National Average Baseline Incidence Rates2
Endpoint
Mortality
Respiratory
Hospital
Admissions.
Asthma ER visits
Minor Restricted
Activity Days
(MRADs)
School Loss
Days
Source
CDC Compressed Mortality File,
accessed through CDC Wonder
(1996-1998)
1999 NHDS public use data files"
2000 NHAMCS public use data
files'; 1999 NHDS public use data
filesb
Ostro and Rothschild
(1989, p. 243)
National Center for Education
Statistics (1996) and 1996 HIS
(Adams et al., 1999, Table 47);
estimate of 180 school days per
year
Notes
non-
accidental
incidence
incidence
incidence
all-cause
Rate per 100 people per yeard by Age Grou]
<18
0.025
0.043
1.011
990.0
18-24
0.022
0.084
1.087
780
25-34
0.057
0.206
0.751
780
35-44
0.150
0.678
0.438
780
45-54
0.383
1.926
0.352
780
3
55-64
1.006
4.389
0.425
780
65+
4.937
11.62
0.232
Notes:
The following abbreviations are used to describe the national surveys conducted by the National Center for Health
Statistics: HIS refers to the National Health Interview Survey; NHDS - National Hospital Discharge Survey; NHAMCS -
National Hospital Ambulatory Medical Care Survey.
b See ftp://ftp.cdc.gov/pub/Health Statistics/NCHS/Datasets/NHDS/
c See ftp://ftp.cdc.gov/pub/Health Statistics/NCHS/Datasets/NHAMCS/
d All of the rates reported here are population-weighted incidence rates per 100 people per year. Additional details on the
incidence and prevalence rates, as well as the sources for these rates are available upon request.
7-104
-------
Environmental and Health Impacts
Table 7-19 National Average Baseline Incidence Rates (continued)
Endpoint
Asthma Exacerbations
Source
Ostro et al. (2001)
Vedaletal. (1998)
Notes
Incidence (and
prevalence)
among
asthmatic
African-
American
children
Incidence (and
prevalence)
among
asthmatic
children
Daily wheeze
Daily cough
Daily dyspnea
Daily wheeze
Daily cough
Daily dyspnea
Rate per 100 people per
year
0.076(0.173)
0.067(0.145)
0.037 (0.074)
0.038
0.086
0.045
7.3.1.3.1.4 Treatment of Potential Thresholds in PM2.5-Related Health Impact Functions
In past analyses, OTAQ has estimated PM2.5-related benefits assuming that a threshold
exists in the PM-related concentration-response functions (at 10 (ag/m3) below which there are no
associations between exposure to PM2.5 and health impacts. Based on our review of the body of
scientific literature, however, EPA's preferred benefits estimation approach assumes a no-
threshold model that calculates incremental benefits down to the lowest modeled PM2.5 air
quality levels.
EPA strives to use the best available science to support our benefits analyses, and we
recognize that interpretation of the science regarding air pollution and health is dynamic and
evolving. EPA's Integrated Science Assessment,QQ which was recently reviewed by EPA's
Clean Air Scientific Advisory Committee,RR'ss concluded that the scientific literature
consistently finds that a no-threshold log-linear model most adequately portrays the PM-
mortality concentration-response relationship while recognizing potential uncertainty about the
QQ U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for Particulate Matter .
National Center for Environmental Assessment, Research Triangle Park, NC. EPA/600/R-08/139F. December.
Available on the Internet at . Accessed March 15,
2010.
RR U.S. Environmental Protection Agency - Science Advisory Board (U.S. EPA-SAB). Review of EPA's
Integrated Science Assessment for Particulate Matter (First External Review Draft, December 2008). EPA-
COUNCIL-09-008. May. Available on the Internet at
.
ss U.S. Environmental Protection Agency - Science Advisory Board (U.S. EPA-SAB). Consultation on EPA's
Particulate Matter National Ambient Air Quality Standards: Scope and Methods Plan for Health Risk and Exposure
Assessment. EPA-COUNCIL-09-009. May. Available on the Internet at
.
105
-------
Regulatory Impact Analysis
exact shape of the concentration-response function.^ Although this document does not
necessarily represent final agency policy, it provides a basis for reconsidering the application of
thresholds in PMi.5 concentration-response functions used in EPA's RIAs. It is important to note
that while CASAC provides advice regarding the science associated with setting the National
Ambient Air Quality Standards, typically other scientific advisory bodies provide specific advice
regarding benefits analysis.uu This approach reflects EPA's most current interpretation of the
scientific literature on PMi.5 and mortality. Please refer to the documentation associated with the
proposed Portland Cement MACT RIA for a description of the history of the treatment of
thresholds in our analyses.324
As can be seen in Table 7-7-20, we conducted a sensitivity analysis for premature
mortality, with alternative thresholds at 3 |^g/m3 (the "background," or no-threshold,
assumption), 7.5 (^g/m3, 10 (^g/m3,12 |ig/m3, and 14 |^g/m3. By replacing the no-threshold
assumption in the ACS premature mortality function with a 10 (^g/m3 threshold model, the
number of avoided incidences of premature mortality would decrease by approximately 22
percent.
TT It is important to note that uncertainty regarding the shape of the concentration-response function is conceptually
distinct from an assumed threshold. An assumed threshold (below which there are no health effects) is a
discontinuity, which is a specific example of non-linearity.
uu In the proposed Portland Cement RIA, EPA solicited comment on the use of the no-threshold model for benefits
analysis within the preamble of that proposed rule. The comment period for the Portland Cement proposed
NESHAP closed on September 4, 2009 (Docket ID No. EPA-HQ-OAR-2002-0051 available at
http://www.regulations.gov). EPA is currently reviewing those comments. U.S. Environmental Protection Agency.
(2009). Regulatory Impact Analysis: National Emission Standards for Hazardous Air Pollutants from the Portland
Cement Manufacturing Industry. Office of Air and Radiation. Retrieved on May 4, 2009, from
http://www.epa.gov/ttn/ecas/regdata/RIAs/portlandcementria_4-20-09.pdf
106
-------
Environmental and Health Impacts
Table 7-7-20 PM-Related Mortality Benefits Associated with the Final Rule: Threshold Sensitivity Analysis
Using the ACS Study (Pope et al., 2002)a
Level of Assumed Threshold
14 ng/m3 b
12 |ag/m3
10 ng/m3 c
7.5 ng/m3 d
3 ng/m3 e
PM Mortality Incidence
2030
30
35
47
56
60
a Note that this table only presents the effects of a threshold on
PM-related mortality incidence based on the ACS study.
b Alternative annual PM NAAQS.
c Previous threshold assumption
dSAB-HES(2004)86
e NAS (2002)87
7.3.1.3.2 Economic Values for Health Outcomes
Reductions in ambient concentrations of air pollution generally lower the risk of future
adverse health effects for a large population. Therefore, the appropriate economic measure is
willingness-to-pay (WTP) for changes in risk of a health effect rather than WTP for a health
effect that would occur with certainty (Freeman, 1993). Epidemiological studies generally
provide estimates of the relative risks of a particular health effect that is avoided because of a
reduction in air pollution. We converted those to units of avoided statistical incidence for ease of
presentation. We calculated the value of avoided statistical incidences by dividing individual
WTP for a risk reduction by the related observed change in risk. For example, suppose a
pollution-reduction regulation is able to reduce the risk of premature mortality from 2 in 10,000
to 1 in 10,000 (a reduction of 1 in 10,000). If individual WTP for this risk reduction is $100, then
the WTP for an avoided statistical premature death is $1 million ($100/0.0001 change in risk).
WTP estimates generally are not available for some health effects, such as hospital
admissions. In these cases, we used the cost of treating or mitigating the effect as a primary
estimate. These cost-of-illness (COI) estimates generally understate the true value of reducing
the risk of a health effect, because they reflect the direct expenditures related to treatment, but
not the value of avoided pain and suffering (Harrington and Portney, 1987; Berger, 1987). We
provide unit values for health endpoints (along with information on the distribution of the unit
value) in Table 7-7-21. All values are in constant year 2000 dollars, adjusted for growth in real
107
-------
Regulatory Impact Analysis
income out to 2020 and 2030 using projections provided by Standard and Poor's. Economic
theory argues that WTP for most goods (such as environmental protection) will increase if real
income increases. Many of the valuation studies used in this analysis were conducted in the late
1980s and early 1990s. Because real income has grown since the studies were conducted,
people's willingness to pay for reductions in the risk of premature death and disease likely has
grown as well. We did not adjust cost of illness-based values because they are based on current
costs. Similarly, we did not adjust the value of school absences, because that value is based on
current wage rates. For details on valuation estimates for PM-related endpoints, see the 2006
PM NAAQS RIA. For details on valuation estimates for ozone-related endpoints, see the 2008
Ozone NAAQS RIA.
108
-------
Environmental and Health Impacts
Table 7-7-21 Unit Values Used for Economic Valuation of Health Endpoints (2000$)a
Health Endpoint
Central Estimate of Value Per Statistical
Incidence
1990 Income
Level
2020 Income
T ib
Level
2030 Income
T 1b
Level
Derivation of Estimates
Premature Mortality (Value of a
Statistical Life): PM2.5- and
Ozone-related
$6,320,000
$7,590,000
$7,800,000
EPA currently recommends a default central VSL of $6.3 million
based on a Weibull distribution fitted to twenty-six published VSL
estimates (5 contingent valuation and 21 labor market studies). The
underlying studies, the distribution parameters, and other useful
information are available in Appendix B of EPA's current Guidelines
for Preparing Economic Analyses. The guidelines can be accessed
at: http://yosemite.epa.gov/ee/epa/eermfile.nsf/vwAN/EE-0516-
01 .pdf/$File/EE-0516-01 .pdf
Chronic Bronchitis (CB)
$340,000
$420,000
$430,000
Point estimate is the mean of a generated distribution of WTP to
avoid a case of pollution-related CB. WTP to avoid a case of
pollution-related CB is derived by adjusting WTP (as described in
Viscusi et al., 1991325) to avoid a severe case of CB for the difference
in severity and taking into account the elasticity of WTP with respect
to severity of CB.
Nonfatal Myocardial Infarction
(heart attack)
3% discount rate
Age 0-24
Age 25^4
Age 45-54
Age 55-65
Age 66 and over
7% discount rate
Age 0-24
Age 25^4
Age 45-54
Age 55-65
Age 66 and over
$66,902
$74,676
$78,834
$140,649
$66,902
$65,293
$73,149
$76,871
$132,214
$65,293
$66,902
$74,676
$78,834
$140,649
$66,902
$65,293
$73,149
$76,871
$132,214
$65,293
$66,902
$74,676
$78,834
$140,649
$66,902
$65,293
$73,149
$76,871
$132,214
$65,293
Age-specific cost-of-illness values reflect lost earnings and direct
medical costs over a 5-year period following a nonfatal MI. Lost
326
earnings estimates are based on Cropper and Krupnick (1990).
Direct medical costs are based on simple average of estimates from
Russell et al. (1998)327 and Wittels et al. (1990).328
Lost earnings:
Cropper and Krupnick (1990). Present discounted value of 5 years
of lost earnings:
at 7%
$7,855
$11,578
$66,920
Direct medical expenses: An average of:
1. Wittels et al. (1990) ($102,658—no discounting)
2. Russell et al. (1998), 5-year period ($22,331 at 3% discount rate;
$21,113 at 7% discount rate)
age of onset: at 3%
25-44 $8,774
45-54 $12,932
55-65 $74,746
109
-------
Regulatory Impact Analysis
Table 7-7-21 Unit Values Used for Economic Valuation of Health Endpoints (2000$)a (continued)
Health Endpoint
Central Estimate of Value Per Statistical
Incidence
1990 Income
Level
2020 Income
T 1b
Level
2030 Income
T 1b
Level
Derivation of Estimates
Hospital Admissions
Chronic Obstructive Pulmonary
Disease (COPD)
(ICD codes 490-492, 494-496)
Pneumonia
(ICD codes 480-487)
Asthma Admissions
All Cardiovascular
(ICD codes 390-429)
Emergency Room Visits for
Asthma
$12,378
$14,693
$6,634
$18,387
$286
$12,378
$14,693
$6,634
$18,387
$286
$12,378
$14,693
$6,634
$18,387
$286
The COI estimates (lost earnings plus direct medical costs) are based
on ICD-9 code-level information (e.g., average hospital care costs,
average length of hospital stay, and weighted share of total COPD
category illnesses) reported in Agency for Healthcare Research and
329
Quality (2000) (www.ahrq.gov).
The COI estimates (lost earnings plus direct medical costs) are based
on ICD-9 code-level information (e.g., average hospital care costs,
average length of hospital stay, and weighted share of total
pneumonia category illnesses) reported in Agency for Healthcare
Research and Quality (2000) (www.ahrq.gov).
The COI estimates (lost earnings plus direct medical costs) are based
on ICD-9 code-level information (e.g., average hospital care costs,
average length of hospital stay, and weighted share of total asthma
category illnesses) reported in Agency for Healthcare Research and
Quality (2000) (www.ahrq.gov).
The COI estimates (lost earnings plus direct medical costs) are based
on ICD-9 code-level information (e.g., average hospital care costs,
average length of hospital stay, and weighted share of total
cardiovascular category illnesses) reported in Agency for Healthcare
Research and Quality (2000) (www.ahrq.gov).
Simple average of two unit COI values:
(1) $311.55, from Smith et al. (1997)330 and
(2) $260.67, from Stanford et al. (1999).331
110
-------
Environmental and Health Impacts
Table 7-7-21 Unit Values Used for Economic Valuation of Health Endpoints (2000$)a (continued)
Health Endpoint
Central Estimate of Value Per Statistical
Incidence
1990 Income
Level
2020 Income
Level
2030 Income
Level
Derivation of Estimates
Respiratory Ailments Not Requiring Hospitalization
Upper Respiratory Symptoms
(URS)
$25
$27
$27
Combinations of the three symptoms for which WTP estimates are
available that closely match those listed by Pope et al. result in seven
different "symptom clusters," each describing a "type" of URS. A
dollar value was derived for each type of URS, using mid-range
332
estimates of WTP (lEc, 1994) to avoid each symptom in the
cluster and assuming additivity of WTPs. The dollar value for URS
is the average of the dollar values for the seven different types of
URS.
Lower Respiratory Symptoms
(LRS)
$16
$17
$17
Combinations of the four symptoms for which WTP estimates are
available that closely match those listed by Schwartz et al. result in
11 different "symptom clusters," each describing a "type" of LRS. A
dollar value was derived for each type of LRS, using mid-range
estimates of WTP lEc, 1994) to avoid each symptom in the cluster
and assuming additivity of WTPs. The dollar value for LRS is the
average of the dollar values for the 11 different types of LRS.
Asthma Exacerbations
$42
$45
$45
Asthma exacerbations are valued at $42 per incidence, based on the
mean of average WTP estimates for the four severity definitions of a
"bad asthma day," described in Rowe and Chestnut (1986). This
study surveyed asthmatics to estimate WTP for avoidance of a "bad
asthma day," as defined by the subjects. For purposes of valuation,
an asthma attack is assumed to be equivalent to a day in which
asthma is moderate or worse as reported in the Rowe and Chestnut
(1986) study.
Acute Bronchitis
$360
$380
$390
Assumes a 6-day episode, with daily value equal to the average of
low and high values for related respiratory symptoms recommended
in Neumann et al. (1994).
Ill
-------
Regulatory Impact Analysis
Table 7-7-21 Unit Values Used for Economic Valuation of Health Endpoints (2000$)a (continued)
Health Endpoint
Central Estimate of Value Per Statistical
Incidence
1990 Income
Level
2020 Income
Level
2030 Income
Level
Derivation of Estimates
Restricted Activity and Work/School Loss Days
Work Loss Days (WLDs)
School Absence Days
Worker Productivity
Minor Restricted Activity Days
(MRADs)
Variable
(national
median = )
$75
$0.95 per
worker per
10% change in
ozone per day
$51
$75
$0.95 per
worker per
10% change
in ozone per
day
$54
$75
$0.95 per
worker per
10% change in
ozone per day
$55
County-specific median annual wages divided by 50 (assuming 2 weeks
of vacation) and then by 5 — to get median daily wage. U.S. Year 2000
Census, compiled by Geolytics, Inc.
Based on expected lost wages from parent staying home with child.
Estimated daily lost wage (if a mother must stay at home with a sick
child) is based on the median weekly wage among women age 25 and
older in 2000 (U.S. Census Bureau, Statistical Abstract of the United
States: 2001, Section 12: Labor Force, Employment, and Earnings,
Table No. 621). This median wage is $551. Dividing by 5 gives an
estimated median daily wage of $103.
The expected loss in wages due to a day of school absence in which the
mother would have to stay home with her child is estimated as the
probability that the mother is in the workforce times the daily wage she
would lose if she missed a day = 72.85% of $103, or $75.
Based on $68 - median daily earnings of workers in farming, forestry
and fishing - from Table 621, Statistical Abstract of the United States
("Full-Time Wage and Salary Workers - Number and Earnings: 1985 to
2000") (Source of data in table: U.S. Bureau of Labor Statistics,
Bulletin 2307 and Employment and Earnings, monthly).
Median WTP estimate to avoid one MR AD from Tolley et al. (1986).
All monetized annual benefit estimates are presented in year 2000 dollars. We use the Consumer Price Indexes to adjust both WTP- and COI-based benefits
estimates to 2007 dollars from 2000 dollars. For WTP-based estimates, we use an inflation factor of 1.20 based on the CPI-U for "all items." For COI-based
estimates, we use an inflation factor of 1.35 based on the CPI-U for medical care.
Our analysis accounts for expected growth in real income over time. Economic theory argues that WTP for most goods (such as environmental protection) will
increase if real incomes increase. Benefits are therefore adjusted by multiplying the unadjusted benefits by the appropriate adjustment factor to account for income
growth over time. For a complete discussion of how these adjustment factors were derived, we refer the reader to the PM NAAQS regulatory impact analysis. Note
that similar adjustments do not exist for cost-of-illness-based unit values. For these, we apply the same unit value regardless of the future year of analysis.
112
-------
Environmental and Health Impacts
7.3.1.3.3 Manipulating Air Quality Modeling Data for Health Impacts Analysis
In Section 7.2, we summarized the methods for and results of estimating air quality for
the final rule. These air quality results are in turn associated with human populations to
estimate changes in health effects. For the purposes of this analysis, we focus on the health
effects that have been linked to ambient changes in ozone and PM2.5 related to emission
reductions estimated to occur due to the implementation of the final rule. We estimate
ambient PM2.5 and ozone concentrations using the Community Multiscale Air Quality model
(CMAQ). This section describes how we converted the CMAQ modeling output into full-
season profiles suitable for the health impacts analysis.
7.3.1.3.3.1 General Methodology
First, we extracted hourly, surface-layer PM and ozone concentrations for each grid
cell from the standard CMAQ output files. For ozone, these model predictions are used in
conjunction with the observed concentrations obtained from the Aerometric Information
Retrieval System (AIRS) to generate ozone concentrations for the entire ozone season.vv>ww
The predicted changes in ozone concentrations from the future-year base case to future-year
control scenario serve as inputs to the health and welfare impact functions of the benefits
analysis (i.e., BenMAP).
To estimate ozone-related health effects for the contiguous United States, full-season
ozone data are required for every BenMAP grid-cell. Given available ozone monitoring data,
we generated full-season ozone profiles for each location in two steps: (1) we combined
monitored observations and modeled ozone predictions to interpolate hourly ozone
concentrations to a grid of 12-km by 12-km population grid cells for the contiguous 48 states,
and (2) we converted these full-season hourly ozone profiles to an ozone measure of interest,
such as the daily 8-hour maximum.**'YY
For PM2.5, we also use the model predictions in conjunction with observed monitor
data. CMAQ generates predictions of hourly PM species concentrations for every grid. The
species include a primary coarse fraction (corresponding to PM in the 2.5 to 10 micron size
range), a primary fine fraction (corresponding to PM less than 2.5 microns in diameter), and
several secondary particles (e.g., sulfates, nitrates, and organics). PM2.5 is calculated as the
sum of the primary fine fraction and all of the secondarily formed particles. Future-year
estimates of PM2.5 were calculated using relative reduction factors (RRFs) applied to 2002
ambient PM2.5 and PM2.5 species concentrations. A gridded field of PM2.5 concentrations was
created by interpolating Federal Reference Monitor ambient data and IMPROVE ambient
vv The ozone season for this analysis is defined as the 5-month period from May to September.
w™ Based on AIRS, there were 961 ozone monitors with sufficient data (i.e., 50 percent or more days reporting
at least nine hourly observations per day [8 am to 8 pm] during the ozone season).
xx The 12-km grid squares contain the population data used in the health benefits analysis model, BenMAP.
YY This approach is a generalization of planar interpolation that is technically referred to as enhanced Voronoi
Neighbor Averaging (EVNA) spatial interpolation. See the BenMAP manual for technical details, available for
download at http://www.epa.gov/air/benmap.
7-113
-------
Regulatory Impact Analysis
data. Gridded fields of PM2.5 species concentrations were created by interpolating EPA
speciation network (ESPN) ambient data and IMPROVE data. The ambient data were
interpolated to the CMAQ 12 km grid.
The procedures for determining the RRFs are similar to those in EPA's draft guidance
for modeling the PM2.5 standard (EPA, 1999). The guidance recommends that model
predictions be used in a relative sense to estimate changes expected to occur in each major
PMi.5 species. The procedure for calculating future-year PM2.5 design values is called the
"Speciated Modeled Attainment Test (SMAT)." EPA used this procedure to estimate the
ambient impacts of the final rule.
Table 7-7-22 provides those ozone and PM^s metrics for grid cells in the modeled
domain that enter the health impact functions for health benefits endpoints. The population-
weighted average reflects the baseline levels and predicted changes for more populated areas
of the nation. This measure better reflects the potential benefits through exposure changes to
these populations.
Table 7-7-22 Summary of CMAQ-Derived Population-Weighted Ozone and PM2.s Air Quality Metrics for
Health Benefits Endpoints Associated with the Final Rule
Statistic2
2030
Baseline
Change"
Ozone Metric: National Population-Weighted Average (ppb)c
Daily Maximum 8 -Hour Average
Concentration
43.5620
-0.0104
PM2.s Metric: National Population- Weighted Average (ug/m3)
Average Concentration
9.7548
0.0036
a Ozone and PM2 5 metrics are calculated at the CMAQ grid-cell level for use in health effects
estimates. Ozone metrics are calculated over relevant time periods during the daylight hours
of the "ozone season" (i.e., May through September).
b The change is defined as the base-case value minus the control-case value. A negative value
indicates an increase in population-weighted average air quality from the baseline to the
control scenario.
c Calculated by summing the product of the projected CMAQ grid-cell population and the
estimated CMAQ grid cell seasonal ozone concentration and then dividing by the total
population.
Emissions and air quality modeling decisions are made early in the analytical process.
For this reason, the emission control scenarios used in the air quality and benefits modeling
are slightly different than the final emission inventories estimated for the final rule. Please
refer to Chapter 5.5 for more information about the inventories used in the air quality
modeling that supports the health impacts analysis.
7.3.1.4 Methods for Describing Uncertainty
The National Research Council (NRC)337 highlighted the need for EPA to conduct
rigorous quantitative analysis of uncertainty in its benefits estimates and to present these
estimates to decision makers in ways that foster an appropriate appreciation of their inherent
7-114
-------
Environmental and Health Impacts
uncertainty. In response to these comments, EPA's Office of Air and Radiation (OAR) is
developing a comprehensive strategy for characterizing the aggregate impact of uncertainty in
key modeling elements on both health incidence and benefits estimates. Components of that
process include emissions modeling, air quality modeling, health effects incidence estimation,
and valuation.
In benefit analyses of air pollution regulations conducted to date, the estimated impact
of reductions in premature mortality has accounted for 85% to 95% of total benefits.
Therefore, it is particularly important to characterize the uncertainties associated with
reductions in premature mortality. The health impact functions used to estimate avoided
premature deaths associated with reductions in ozone have associated standard errors that
represent the statistical errors around the effect estimates in the underlying epidemiological
studies.22 In our results, we report credible intervals based on these standard errors, reflecting
the uncertainty in the estimated change in incidence of avoided premature deaths. We also
provide multiple estimates, to reflect model uncertainty between alternative study designs.
For premature mortality associated with exposure to PM, we follow the same approach
that has been used in several recent RjAs.338'339'340 First, we use Monte Carlo methods for
estimating random sampling error associated with the concentration response functions from
epidemiological studies and economic valuation functions. Monte Carlo simulation uses
random sampling from distributions of parameters to characterize the effects of uncertainty on
output variables, such as incidence of premature mortality. Specifically, we used Monte
Carlo methods to generate confidence intervals around the estimated health impact and dollar
benefits. Distributions for individual effect estimates are based on the reported standard
errors in the epidemiological studies. Distributions for unit values are described in Table 7-
7-21.
Second, as a sensitivity analysis, we use the results of our expert elicitation of the
concentration response function describing the relationship between premature mortality and
ambient PM2.5 concentration.AAA'341 Incorporating only the uncertainty from random
sampling error omits important sources of uncertainty (e.g., in the functional form of the
model; whether or not a threshold may exist). This second approach attempts to incorporate
these other sources of uncertainty.
Use of the expert elicitation and incorporation of the standard errors approaches
provide insights into the likelihood of different outcomes and about the state of knowledge
regarding the benefits estimates. Both approaches have different strengths and weaknesses,
which are fully described in Chapter 5 of the PM NAAQS RIA.
These multiple characterizations, including confidence intervals, omit the contribution
22 Health impact functions measure the change in a health endpoint of interest, such as hospital admissions, for a
given change in ambient ozone or PM concentration.
AAA Expert elicitation is a formal, highly structured and well documented process whereby expert judgments,
usually of multiple experts, are obtained (Ayyb, 2002).
7-115
-------
Regulatory Impact Analysis
to overall uncertainty of uncertainty in air quality changes, baseline incidence rates,
populations exposed and transferability of the effect estimate to diverse locations.
Furthermore, the approach presented here does not yet include methods for addressing
correlation between input parameters and the identification of reasonable upper and lower
bounds for input distributions characterizing uncertainty in additional model elements. As a
result, the reported confidence intervals and range of estimates give an incomplete picture
about the overall uncertainty in the estimates. This information should be interpreted within
the context of the larger uncertainty surrounding the entire analysis.
As mentioned above, total benefits are driven primarily by the reduction in PM2.5-
related premature mortalities each year. Some key assumptions underlying the premature
mortality estimates include the following, which may also contribute to uncertainty:
• Inhalation of fine particles is causally associated with premature death at
concentrations near those experienced by most Americans on a daily basis.
Although biological mechanisms for this effect have not yet been completely
established, the weight of the available epidemiological, lexicological, and
experimental evidence supports an assumption of causality. The impacts of
including a probabilistic representation of causality were explored in the expert
elicitation-based results of the PM NAAQS RIA.
• All fine particles, regardless of their chemical composition, are equally potent
in causing premature mortality. This is an important assumption, because PM
produced via transported precursors emitted from engines may differ
significantly from PM precursors released from electric generating units and
other industrial sources. However, no clear scientific grounds exist for
supporting differential effects estimates by particle type.
• The C-R function for fine particles is approximately linear within the range of
ambient concentrations under consideration. Thus, the estimates include health
benefits from reducing fine particles in areas with varied concentrations of PM,
including both regions that may be in attainment with PM2.5 standards and
those that are at risk of not meeting the standards.
• There is uncertainty in the magnitude of the association between ozone and
premature mortality. The range of ozone impacts associated with the final rule
is estimated based on the risk of several sources of ozone-related mortality
effect estimates. In a recent report on the estimation of ozone-related
premature mortality published by the National Research Council, a panel of
experts and reviewers concluded that short-term exposure to ambient ozone is
likely to contribute to premature deaths and that ozone-related mortality should
be included in estimates of the health benefits of reducing ozone exposure.342
EPA has requested advice from the National Academy of Sciences on how best
to quantify uncertainty in the relationship between ozone exposure and
premature mortality in the context of quantifying benefits.
7-116
-------
Environmental and Health Impacts
Acknowledging omissions and uncertainties, we present a best estimate of the total
benefits based on our interpretation of the best available scientific literature and methods
supported by EPA's technical peer review panel, the Science Advisory Board's Health Effects
Subcommittee (SAB-HES). The National Academies of Science (NRC, 2002) has also
reviewed EPA's methodology for analyzing the health benefits of measures taken to reduce
air pollution. EPA addressed many of these comments in the analysis of the final PM
NAAQS.343'344 This analysis incorporates this most recent work to the extent possible.
7.3.2 PM-related Monetized Benefits of the Model Year (MY) Analysis
As described in Chapter 5.5, the final standards will reduce emissions of several
criteria and toxic pollutants and precursors. In the MY analysis, EPA estimates the economic
value of the human health benefits associated with reducing PM2.5 exposure. Due to
analytical limitations, this analysis does not estimate benefits related to other criteria
pollutants (such as ozone, NO2 or SO2) or toxics pollutants, nor does it monetize all of the
potential health and welfare effects associated with PM2.5.
The MY analysis uses a "benefit-per-ton" method to estimate a selected suite of
related health benefits described below. These PM2.5 benefit-per-ton estimates provide the
total monetized human health benefits (the sum of premature mortality and premature
morbidity) of reducing one ton of directly emitted PM2.5, or its precursors (such as NOx, SOx,
and VOCs), from a specified source. Ideally, the human health benefits associated with the
MY analysis would be estimated based on changes in ambient PM2.5 as determined by full-
scale air quality modeling. However, this modeling was not possible in the timeframe for the
final rule due to the time and resource constraints associated with running full-scale
photochemical air quality modeling.
The dollar-per-ton estimates used in this analysis are provided in Table 7-23. In the
summary of costs and benefits, Chapter 8.4 of this RIA, EPA presents the monetized value of
PM-related improvements associated with the final rule.
7-117
-------
Regulatory Impact Analysis
Table 7-23 Benefits-per-ton Values (2007$) Derived Using the ACS Cohort Study for PM-related
Premature Mortality (Pope et al., 2002)a
Yearc
All Sources'1
SOX
VOC
Stationary (Non-EGU)
Sources
NOX
Direct PM2.5
Mobile Sources
NOX
Direct PM2.5
Estimated Using a 3 Percent Discount Rate"
2015
2020
2030
2040
$28,000
$31,000
$36,000
$43,000
$1,200
$1,300
$1,500
$1,800
$4,700
$5,100
$6,100
$7,200
$220,000
$240,000
$280,000
$330,000
$4,900
$5,300
$6,400
$7,600
$270,000
$290,000
$350,000
$420,000
Estimated Using a 7 Percent Discount Rate"
2015
2020
2030
2040
$26,000
$28,000
$33,000
$39,000
$1,100
$1,200
$1,400
$1,600
$4,200
$4,600
$5,500
$6,600
$200,000
$220,000
$250,000
$300,000
$4,400
$4,800
$5,800
$6,900
$240,000
$270,000
$320,000
$380,000
aThe benefit-per-ton estimates presented in this table are based on an estimate of premature mortality derived
from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six-Cities study
(Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
b The benefit-per-ton estimates presented in this table assume either a 3 percent or 7 percent discount rate in the
valuation of premature mortality to account for a twenty-year segmented cessation lag.
c Benefit-per-ton values were estimated for the years 2015, 2020, and 2030. For 2040, EPA 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,345 the proposed
Portland Cement National Emissions Standards for Hazardous Air Pollutants (NESHAP) RIA
346
347
(U.S. EPA, 2009a), and the final NO2 NAAQS (U.S. EPA, 2009b). Table 7-24 shows
the quantified and unquantified PM2.5-related co-benefits captured in those benefit-per-ton
estimates.
Table 7-24 Human Health and Welfare Effects of PM2.5
Pollutant /
Effect
Quantified and Monetized
in Primary Estimates
Unquantified Effects
Changes in:
PM2.5
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
7-118
-------
Environmental and Health Impacts
Consistent with the final NO2 NAAQS,BBB 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)34 accompanying the
recent final ozone NAAQS RIA. Readers can also refer to Fann et al. (2009)349 for a detailed
description of the benefit-per-ton methodology.ccc 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
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
BBB Although we summarize the main issues in this chapter, we encourage interested readers to see the benefits
chapter of the final primary NO2 NAAQS RIA for a more detailed description of recent changes to the PM
benefits presentation and preference for the no-threshold model.
ccc 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
7-119
-------
Regulatory Impact Analysis
mortality that EPA usually conducts in association with its benefits analysis was not
conducted for this analysis.
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)350 and the Harvard Six-Cities
cohort (Laden et al., 2006).351 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).352'353
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
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)354 meta-analysis of 33 studies. The $10 million value represented the upper
end of the interquartile range from the Viscusi and Aldy (2003)355 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)356 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
7-120
-------
Environmental and Health Impacts
for Preparing Economic Analyses (U.S. EPA, 2000)357 while they continue efforts to update
their guidance on this issue.0 D 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.
• Dollar-per-ton estimates do not reflect local variability in population density,
meteorology, exposure, baseline health incidence rates, or other local factors
that might lead to an overestimate or underestimate of the actual benefits of
controlling fine particulates. In Chapter 7.2, we describe the full-scale air
quality modeling conducted for the 2030 calendar year analysis in an effort to
capture this variability.
• There are several health benefits categories that EPA was unable to quantify in
the MY analysis due to limitations associated with using benefits-per-ton
estimates, several of which could be substantial. Because NOx and VOC
emissions are also precursors to ozone, changes in NOx and VOC would also
impact ozone formation and the health effects associated with ozone exposure.
Benefits-per-ton estimates for ozone, however, 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 the previous section of this RIA (Chapter 7.3.1)for a description
of the quantification and monetization of health impacts for the CY analysis
and a description of the unquantified co-pollutant benefits associated with this
rulemaking.
• The benefit-per-ton estimates used in this analysis incorporate projections of
key variables, including atmospheric conditions, source level emissions,
population, health baselines and incomes, technology. These projections
introduce some uncertainties to the benefit per ton estimates.
• As described above, using the benefit-per-ton value derived from the ACS
study (Pope et al., 2002) alone provides an incomplete characterization of
PM2.5 benefits. When placed in the context of the Expert Elicitation results,
this estimate falls toward the lower end of the distribution. By contrast, the
estimated PMi.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.
DDD 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 .
7-121
-------
Regulatory Impact Analysis
As mentioned above, emissions changes and benefits-per-ton estimates alone are not a
good indication of local or regional air quality and health impacts, as there may be localized
impacts associated with this ralemaking. Additionally, the atmospheric chemistry related to
ambient concentrations of PM2.5, ozone and air toxics is very complex. Full-scale
photochemical modeling is therefore necessary to provide the needed spatial and temporal
detail to more completely and accurately estimate the changes in ambient levels of these
pollutants and their associated health and welfare impacts. Timing and resource constraints
precluded EPA from conducting full-scale photochemical air quality modeling for the MY
analysis. We have, however, conducted national-scale air quality modeling for the CY
analysis to analyze the impacts of the standards on PM2.5, ozone, and selected air toxics.
7.4 Changes in Global Mean Temperature and Sea Level Rise Associated
with the Rule'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 rule will affect climate change projections. Since GHGs are well-
mixed 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 provides estimates for the response in global mean surface
temperature and sea level rise projections to the estimated net global GHG emissions
reductions associated with the this rule (see Chapter 5 for the estimated net reductions in
global emissions over time by GHG).
7.4.2 Estimated Projected Reductions in Atmospheric COa Concentrations,
Global Mean Surface Temperatures and Sea Level Rise
To assess the impact of the emissions reductions from the rule, EPA estimated
changes in projected atmospheric CC>2 concentrations, global mean surface temperature and
sea-level rise to 2100 using the GCAM (Global Change Assessment Model, formerly
MiniCAM), integrated assessment modelEEE coupled with the MAGICC (Model for the
EEE
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
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.
7-122
-------
Environmental and Health Impacts
Assessment of Greenhouse-gas Induced Climate Change) simple climate model.FFF GCAM
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 COi 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 rule was developed by applying the rule's estimated
emissions reductions to the GCAM reference (no climate policy or baseline) scenario (used as
C^C^C^
the basis for the Representative Concentration Pathway RCP4.5 ). Specifically, the CC>2,
N2O, CH4, and HFC-134a emissions reductions from Chapter 5 were applied as net reductions
to the GCAM global baseline net emissions for each GHG. All emissions reductions were
assumed to begin in 2012, with zero emissions change in 2011 and linearly increasing to
equal the value supplied (in Chapter 5) for 2020. The emissions reductions past 2050 were
FFF 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, MFCs, 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.
GGGThis 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, Nebojsa 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) 132pp.
Smith, Steven J. and T.M.L. Wigley (2006) "Multi-Gas Forcing Stabilization with the MiniCAM" Energy
Journal (Special Issue #3).
7-123
-------
Regulatory Impact Analysis
scaled with total U.S. road transportation fuel consumption from the GCAM reference
scenario. Using MAGICC, the change in atmospheric COi concentrations, global mean
temperature, and sea level were projected at five-year time steps to 2100 for both the
reference (no climate policy) scenario and the emissions reduction scenario specific to this
action. To capture some of the uncertainty in the climate system, the changes in projected
atmospheric CO2 concentrations, global mean temperature 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.HHH
To compute the reductions in atmospheric COi concentration, temperature, and sea
level rise specifically attributable to the rule, the output from the rule's emissions scenario
was subtracted from the reference (no policy or baseline) emissions case scenario. As a result
of the rule's specified emissions reductions, the atmospheric COi concentration is projected to
be reduced by approximately 2.7 to 3.1 parts per million (ppm), the global mean temperature
is projected to be reduced by approximately 0.006-0.015°C by 2100 and global mean sea level
rise is projected to be reduced by approximately 0.06-0.14 cm by 2100.
Figure 7-25 provides the results for the estimated reductions in atmospheric CO2
concentration associated with the rule. Figure 7-26 provides the estimated reductions in
projected global mean temperatures associated with the rule. Figure 7-27 provides the
estimated reductions in global mean sea level rise associated with the rule.
HHH 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/.
7-124
-------
Environmental and Health Impacts
2000
0.0
-0.5
-? -1.0
g
a.
& -1.5
8
I -2.0
i
a -2.5
CO2 Concentration Change
2020
2040
2060
2080
2100
ClimSens=1.5
-ClimSens=2.0
ClimSens=2.5
ClimSens=3.0
-ClimSens=4.5
ClimSens=6.0
Year
Figure 7-25 Estimated Projected Reductions in Atmospheric CO2 Concentrations (parts per million by
volume) from Baseline for the Final Vehicles Rulemaking (for climate sensitivities ranging from 1.5-6°C)
Global Mean Temperature Change
2000
0.0000 -i
-0.0020 -
-0.0040 -
-0.0060 -
-0.0080 -
-0.0100 -
-0.0120 -
-0.0140 -
-0.0160 -
-0.0180
2020
2040
2060
2080
2100
ClimSens=1.5
-ClimSens=2.0
ClimSens=2.5
ClimSens=3.0
-ClimSens=4.5
ClimSens=6.0
Year
Figure 7-26 Estimated Projected Reductions in Global Mean Surface Temperatures from Baseline for the
Final Vehicles Rulemaking (for climate sensitivities ranging from 1.5-6°C)
7-125
-------
Regulatory Impact Analysis
Global Mean Sea level Rise
2000
0.00 -i
2020
2040
2060
2080
2100
-0.02 -
-0.04 -
-0.06 -
-0.08 -
-0.10 -
-0.12 -
-0.14 -
-0.16
ClimSens=1.5
-ClimSens=2.0
ClimSens=2.5
ClimSens=3.0
-ClimSens=4.5
ClimSens=6.0
Year
Figure 7-27 Estimated Projected Reductions in Global Mean Sea Level Rise from Baseline for the Final
Vehicles Rulemaking (for climate sensitivities ranging from 1.5-6°C)
The results in both Figure 7-26 and Figure 7-27 show a relatively small reduction in
the projected global mean 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.
358
In today's rule, EPA analyzes another climate-related variable and calculates the
projected changes in tropical ocean pH. EPA estimated the change in ocean pH using the
Program CO2SYS,ni version 1.05, a program which performs calculations relating parameters
of the carbon dioxide (COi) system in seawater. EPA used the program to calculate ocean pH
as a function of atmospheric CC>2, among other specified input conditions. Based on the
projected atmospheric COi concentration reductions (average of 2.9 ppm by 2100) that would
result from this rule, the program calculates an increase in ocean pH of approximately 0.0014
pH units in 2100. Thus, this analysis indicates the projected decrease in atmospheric CC>2
concentrations from today's rule would result in an increase in ocean pH.
7.4.3 Summary
EPA's analysis of the rule's effect on global climate conditions is intended to quantify
these potential reductions using the best available science. While EPA's modeling results of
m Lewis, E., and D. W. R. Wallace. 1998. Program Developed for CO2 System Calculations. ORNL/CDIAC-
105. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy,
Oak Ridge, Tennessee.
7-126
-------
Environmental and Health Impacts
the effect of this rale alone show small differences in climate effects (COi concentration,
temperature, sea-level rise, ocean pH), when expressed in terms of global climate endpoints
and global GHG emissions, they yield results that are repeatable and consistent within the
modeling frameworks used.
These projected reductions are proportionally representative of changes to U.S. GHG
emissions in the transportation sector. While not formally estimated for this ralemaking, a
reduction in projected global mean temperature and sea level rise implies a reduction in the
risks associated with 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 (see Chapter 7.5). There are substantial uncertainties in modeling the
global risks of climate change, which complicates quantification and cost-benefits
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., forest disturbance
and catastrophic events such as collapse of large ice sheets and subsequent sea level rise).
7.5 SCC and GHG Benefits
We assigned a monetary value to reductions in CC>2 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 COi emissions, 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 CC>2
emissions, procedures used to model the economic impacts of climate change, and the choice
of discount rates. Furthermore, 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."359
In today's final rale, EPA assigned a dollar value to reductions in CO2 emissions using
SCC estimates that were recently developed by an interagency process. The general approach
to estimating SCC values was to run three integrated assessment models (FUND, DICE, and
PAGE) using inputs agreed upon by the interagency group. The technical support document,
Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866, (i.e.,
SCC TSD) presents a more detailed description of the methodology used to generate the new
estimates, the underlying assumptions, and the limitations of the new SCC estimates.360
For 2010, these estimates are $5, $21, $35, and $65 (in 2007 dollars), and are based on
a COi emissions change of 1 metric ton in 2010. The first three estimates are based on the
average SCC across models and socio-economic and emissions scenarios at the 5, 3, and 2.5
percent discount rates, respectively. The fourth value is included to represent the higher-than -
expected impacts from temperature change further out in the tails of the SCC distribution. For
7-127
-------
Regulatory Impact Analysis
this purpose, EPA has used the SCC value for the 95th percentile at a 3 percent discount rate.
The central value is the average SCC across models at the 3 percent discount rate.
For purposes of capturing the uncertainties involved in regulatory impact analysis, we
emphasize the importance and value of considering the full range. These SCC estimates also
grow over time. For instance, the central value increases to $24 per ton of COi in 2015 and
$26 per ton of CCh in 2020. The table below shows how the SCC estimates change between
2010 and 2050. The complete model results are available in the docket for this final rule
[EPA-HQ-OAR-2009-0472].
Table 7-25 Social Cost of CO2,2010 - 2050 (in 2007 dollars)
Discount Rate
Year
2010
2020
2030
2040
2050
5%
Avg
5
7
10
13
16
3%
Avg
21
26
33
39
45
2.5%
Avg
35
42
50
58
65
3%
95th
65
81
100
119
136
The tables below summarize the total GHG benefits for the lifetime of the rule, which
are calculated by using the four new SCC values. Specifically, total monetized benefits in
each year are calculated by multiplying the marginal benefits estimates per metric ton of CC>2
(the SCC) by the reductions in CO2 for that year. However, these monetized GHG benefits
exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4, NiO) expected under
this final rule. Although EPA has not monetized the benefits of reductions in non-COi GHGs,
the value of these reductions should not be interpreted as zero. Rather, the reductions in non-
CO2 GHGs will contribute to this rule's climate benefits, as explained in Section III.F. The
SCC TSD notes the difference between the social cost of non-CO2 emissions and SCC and
specifies a goal to develop methods to value non-CCh emissions in future analyses.
7-128
-------
Environmental and Health Impacts
Table 7-26 Upstream and Downstream CO2 Benefits for the Given SCC Value, Calendar Year Analysis3 (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
NPVb
5%
(AVERAGE SCC)
$5 IN 2010
$31
$78
$143
$239
$373
$512
$651
$797
$946
$1,111
$1,277
$1,449
$1,621
$1,799
$1,973
$2,151
$2,321
$2,494
$2,664
$2,839
$3,012
$3,191
$3,372
$3,561
$3,752
$3,951
$4,152
$4,362
$4,577
$4,802
$5,033
$5,274
$5,523
$5,781
$6,046
$6,322
$6,606
$6,899
$7,202
$34,500
3%
(AVERAGE SCC)
$21 IN 2010
$135
$334
$602
$989
$1,525
$2,060
$2,586
$3,120
$3,657
$4,221
$4,768
$5,322
$5,868
$6,418
$6,947
$7,475
$7,973
$8,468
$8,949
$9,438
$9,915
$10,408
$10,900
$11,412
$11,926
$12,460
$12,998
$13,558
$14,127
$14,690
$15,267
$15,866
$16,481
$17,118
$17,773
$18,449
$19,143
$19,859
$20,596
$176,700
2.5%
(AVERAGE SCC)
$35 IN 2010
$220
$541
$972
$1,593
$2,447
$3,295
$4,123
$4,960
$5,799
$6,663
$7,496
$8,333
$9,151
$9,972
$10,756
$11,534
$12,261
$12,981
$13,677
$14,381
$15,065
$15,770
$16,472
$17,202
$17,932
$18,689
$19,451
$20,243
$21,046
$21,815
$22,601
$23,417
$24,253
$25,117
$26,004
$26,919
$27,857
$28,823
$29,816
$299,600
3%
(95™ PERCENTILE)
$65 IN 2010
$412
$1,016
$1,834
$3,020
$4,660
$6,301
$7,918
$9,563
$11,223
$12,944
$14,615
$16,303
$17,966
$19,639
$21,250
$22,855
$24,365
$25,867
$27,326
$28,807
$30,252
$31,744
$33,234
$34,785
$36,339
$37,954
$39,582
$41,274
$42,992
$44,691
$46,428
$48,234
$50,087
$52,005
$53,976
$56,013
$58,103
$60,258
$62,476
$538,500
a Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O)
expected under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2
emissions, the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2
GHGs will contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the
difference between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop
methods to value non-CO2 emissions in future analyses.
7-129
-------
Regulatory Impact Analysis
b
Note that net present value of reduced GHG emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used to
calculate net present value of SCC for internal consistency. Refer to SCC TSD for more detail.
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
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 RIA chapter 5. The GHG benefits of the full life of each of the five model years
from 2012 through 2016 are shown in Table 7-27 through Table 7-30 for each of the four
different social cost of carbon values. The GHG benefits are shown for each year in the
model year life and in net present value. The same discount rate used to discount the value of
damages from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present
value of SCC for internal consistency.
7-130
-------
Environmental and Health Impacts
Table 7-27 Upstream and Downstream CO2 Benefits for the 5% (Average SCC) Value, Model Year Analysis3
(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,
5%
MY 2012
$36
$36
$36
$36
$36
$35
$34
$33
$31
$30
$29
$26
$23
$20
$17
$14
$12
$10
$8
$7
$6
$5
$4
$3
$3
$3
$2
$2
$1
$1
$1
$1
$1
$1
$1
$1
$0
$0
$0
$400
MY 2013
$0
$53
$54
$54
$54
$53
$52
$51
$49
$47
$45
$43
$39
$35
$30
$25
$21
$18
$15
$12
$10
$8
$7
$6
$5
$4
$4
$3
$3
$2
$2
$2
$2
$1
$1
$1
$1
$0
$0
$500
MY 2014
$0
$0
$73
$74
$74
$74
$73
$71
$69
$67
$65
$62
$59
$54
$47
$41
$34
$29
$24
$20
$17
$14
$11
$10
$8
$7
$6
$5
$4
$4
$3
$3
$2
$2
$2
$2
$2
$1
$0
$700
MY 2015
$0
$0
$0
$107
$108
$108
$107
$106
$104
$102
$99
$95
$91
$86
$79
$69
$59
$50
$42
$35
$29
$24
$20
$16
$14
$12
$10
$9
$7
$5
$5
$4
$4
$3
$3
$3
$2
$2
$2
$1,000
MY 2016
$0
$0
$0
$0
$151
$151
$152
$150
$148
$147
$144
$140
$134
$128
$120
$111
$97
$83
$70
$58
$49
$40
$33
$27
$23
$19
$16
$14
$12
$10
$8
$7
$6
$5
$5
$4
$4
$3
$3
$1,300
SUM
$36
$89
$163
$271
$422
$421
$418
$411
$402
$393
$382
$366
$346
$322
$293
$259
$223
$189
$158
$132
$110
$91
$76
$63
$53
$45
$38
$32
$27
$23
$19
$17
$15
$13
$12
$10
$9
$7
$5
$3,800
a As noted above, these monetized GHG benefits exclude the value of reductions in non-CO2 GHG
emissions expected under this final rule. Although EPA has not monetized the benefits of reductions in
non-CO2 GHGs, the value of these reductions should not be interpreted as zero.
7-131
-------
Regulatory Impact Analysis
Table 7-28 Upstream and Downstream CO2 Benefits for the 3% (Average SCC) SCC Value, Model Year Analysis3
(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%
MY 2012
$155
$154
$152
$149
$146
$141
$136
$129
$122
$115
$107
$97
$84
$71
$59
$49
$40
$33
$27
$22
$18
$15
$13
$11
$9
$8
$6
$5
$5
$4
$4
$3
$3
$2
$2
$2
$0
$0
$0
$1,700
MY 2013
$0
$230
$227
$225
$220
$215
$208
$199
$189
$179
$169
$157
$142
$124
$105
$88
$73
$60
$50
$41
$34
$28
$23
$19
$17
$14
$12
$9
$8
$7
$6
$6
$5
$4
$4
$3
$3
$0
$0
$2,400
MY 2014
$0
$0
$310
$306
$303
$296
$289
$279
$268
$256
$242
$228
$212
$192
$167
$141
$118
$98
$81
$67
$55
$45
$37
$31
$26
$22
$19
$16
$12
$11
$9
$8
$7
$7
$6
$5
$4
$4
$0
$3,100
MY 2015
$0
$0
$0
$445
$439
$435
$425
$415
$401
$387
$369
$350
$329
$306
$277
$240
$203
$169
$140
$116
$95
$78
$64
$52
$44
$37
$31
$26
$23
$17
$15
$13
$12
$10
$9
$8
$7
$6
$5
$4,400
MY 2016
$0
$0
$0
$0
$616
$608
$601
$589
$574
$557
$537
$513
$485
$457
$424
$384
$332
$281
$234
$194
$160
$131
$107
$88
$73
$61
$51
$43
$37
$31
$23
$20
$18
$16
$14
$12
$11
$10
$8
$5,900
SUM
$155
$383
$689
$1,124
$1,723
$1,695
$1,659
$1,611
$1,554
$1,494
$1,425
$1,345
$1,253
$1,149
$1,032
$902
$766
$641
$532
$439
$362
$297
$244
$202
$168
$142
$119
$100
$84
$70
$57
$51
$45
$39
$35
$31
$25
$20
$14
$17,000
a As noted above, these monetized GHG benefits exclude the value of reductions in non-CO2 GHG
emissions expected under this final rule. Although EPA has not monetized the benefits of reductions in
non-CO2 GHGs, the value of these reductions should not be interpreted as zero.
7-132
-------
Environmental and Health Impacts
Table 7-29 Upstream and Downstream CO2 Benefits for the from 2.5% (Average SCC) SCC Value, Model Year
Analysis3 (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,
2.5%
MY 2012
$253
$249
$246
$241
$234
$226
$216
$205
$193
$181
$168
$152
$131
$110
$92
$76
$62
$51
$41
$34
$28
$23
$19
$16
$14
$12
$9
$8
$7
$6
$5
$5
$4
$4
$3
$3
$0
$0
$0
$2,700
MY 2013
$0
$372
$366
$362
$353
$344
$331
$317
$300
$283
$266
$246
$222
$192
$162
$135
$112
$93
$76
$62
$51
$42
$35
$29
$25
$21
$18
$14
$12
$11
$9
$8
$7
$6
$6
$5
$4
$0
$0
$3,900
MY 2014
$0
$0
$500
$492
$486
$474
$461
$444
$425
$404
$381
$357
$331
$299
$258
$218
$182
$150
$124
$102
$83
$68
$56
$47
$39
$34
$28
$24
$18
$16
$14
$12
$11
$10
$8
$7
$6
$6
$0
$5,200
MY 2015
$0
$0
$0
$716
$705
$695
$679
$660
$636
$611
$581
$547
$514
$475
$429
$370
$312
$259
$214
$176
$144
$118
$96
$79
$66
$55
$47
$39
$34
$25
$22
$19
$17
$15
$13
$12
$10
$9
$8
$7,200
MY 2016
$0
$0
$0
$0
$988
$972
$959
$936
$910
$880
$845
$803
$757
$710
$656
$593
$511
$430
$358
$295
$243
$199
$162
$133
$109
$91
$76
$65
$54
$47
$34
$30
$27
$23
$21
$18
$16
$14
$12
$9,700
SUM
$253
$621
$1,113
$1,810
$2,766
$2,711
$2,646
$2,562
$2,463
$2,359
$2,240
$2,106
$1,954
$1,786
$1,597
$1,392
$1,178
$983
$813
$670
$550
$450
$369
$304
$253
$213
$178
$150
$125
$104
$85
$75
$66
$58
$51
$45
$37
$29
$20
$29,000
a As noted above, these monetized GHG benefits exclude the value of reductions in non-CO2 GHG
emissions expected under this final rule. Although EPA has not monetized the benefits of reductions in
non-CO2 GHGs, the value of these reductions should not be interpreted as zero.
7-133
-------
Regulatory Impact Analysis
Table 7-30 Upstream and Downstream CO2 Benefits for the 3% (95th Percentile) SCC Value, Model Year Analysis3
(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%
MY 2012
$473
$468
$465
$456
$446
$432
$415
$395
$373
$352
$327
$297
$257
$217
$181
$150
$124
$101
$83
$68
$56
$47
$39
$33
$28
$24
$18
$16
$14
$12
$11
$10
$8
$7
$6
$6
$0
$0
$0
$5,100
MY 2013
$0
$699
$691
$685
$672
$657
$636
$611
$581
$550
$518
$481
$436
$378
$320
$268
$222
$184
$152
$124
$102
$85
$71
$59
$51
$43
$37
$28
$25
$22
$19
$17
$15
$13
$12
$10
$9
$0
$0
$7,300
MY 2014
$0
$0
$944
$933
$925
$907
$886
$856
$822
$785
$743
$699
$649
$589
$510
$432
$361
$299
$248
$204
$167
$137
$113
$95
$79
$68
$58
$50
$37
$33
$29
$25
$22
$20
$17
$15
$14
$12
$0
$9,600
MY 2015
$0
$0
$0
$1,357
$1,342
$1,329
$1,303
$1,272
$1,230
$1,186
$1,132
$1,071
$1,008
$936
$848
$733
$619
$517
$428
$353
$290
$237
$194
$160
$134
$112
$96
$81
$69
$51
$45
$40
$35
$31
$27
$24
$21
$19
$16
$13,000
MY 2016
$0
$0
$0
$0
$1,882
$1,859
$1,841
$1,804
$1,761
$1,709
$1,647
$1,572
$1,486
$1,398
$1,297
$1,175
$1,016
$858
$715
$592
$489
$401
$328
$268
$221
$185
$154
$132
$111
$95
$70
$62
$55
$48
$43
$38
$33
$29
$26
$18,000
SUM
$473
$1,167
$2,100
$3,432
$5,267
$5,184
$5,081
$4,939
$4,767
$4,582
$4,367
$4,120
$3,836
$3,517
$3,156
$2,757
$2,342
$1,959
$1,625
$1,341
$1,104
$906
$745
$616
$513
$432
$362
$305
$256
$213
$174
$154
$136
$120
$106
$93
$77
$60
$42
$53,000
a As noted above, these monetized GHG benefits exclude the value of reductions in non-CO2 GHG
emissions expected under this final rule. Although EPA has not monetized the benefits of reductions in
non-CO2 GHGs, the value of these reductions should not be interpreted as zero.
7-134
-------
Environmental and Health Impacts
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 7-28 below from EPA's Fuel Economy Trends Report).361 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. As the figure also shows, little of this efficiency improvement has
been realized in fuel economy gains or GHG emissions reductions.
Fuel Economy Performance
(Three Year Moving Average}
Cars
PC.. 0 to 60 isee.i I'nsrftj Utttttff ,ffisj f
10
25 ..-, 40QO
Adjusted Composite MP
IS--.
0 to 60 Tim* ""•-.,.
- - JMO
" -
. . I in. I i i . . I .... I ii . . I . ... I i i I 2000
1975 I960 1665 1«0 1MS MOO 2003
Year
Fuel Economy Performance'
(Three Year Moving Average)
Wagons
&, oto60t.se>:.}
-•4500
Adjusted Composite MPC.
25- .»,. :_2T_ _____'______ 4000
- • SC'OQ
o to w Tinw "•-...
1985 1MO 19M 2000 200S
Model Year
Figure 7-28 Weight, O-to-60 MPH acceleration time and adjusted fuel economy for light-duty vehicles
During this same period, due in part to increasing numbers of Federal Motor Vehicle
Safety Standards and increasingly stringent NCAP standards from NHTSA, 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 anti-lock brakes (ABS), electronic stability control (ESC), and
improved tires and suspensions help drivers avoid accidents. NHTSA stated in its NPRM that
it anticipated a 12.6 percent reduction in fatality levels between 2007 and 2020 with safety
improvements due to pending NHTSA FMVSS and other factors, such as behavioral
improvements (less drunk driving and increased seat belt usage, for example).JJJ'KKK
jjj.
NHTSA stated in section IX of their PRIA for this rule: "The agency examined the impacts of identifiable
safety trends over the lifetime of the vehicles produced in each model year. An estimate of these impacts was
7-135
-------
Regulatory Impact Analysis
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 COi emissions, the reduction of vehicle
weight is a controversial and complicated topic. The EPA believes that though much work
has been done to advance the understanding of the effects of vehicle weight on safety, there is
much yet to do. EPA also acknowledges that for the analysis of various topics contained in
this RIA and in the preamble, there are a number of uncertainties which exist with these
analysis and any results must be viewed in light of those uncertainties. In the context of
safety, for example, due to limitations in modeling consumer and producer behavioral
responses to the new standards, this analysis does not explore the net fleetwide safety
implications of potential changes in the distribution of vehicles in the on road fleet (e.g.,
changes in the variance of size, weight, vehicle types on the road) if consumers respond to the
rulemaking by purchasing a distribution of vehicles which are significantly different than
what EPA has forecast.
This section of the RIA describes and compares some of the key studies that have been
conducted in the recent past - including NHTSA's recent 2010 analysis, and presents some
potential research options going forward.
7.6.1 What did EPA say in the NPRM and in the Draft RIA with regard to
Potential Safety Effects?
In the NPRM, EPA discussed potential safety effects of the proposed standards. In the
joint technology analysis, EPA and NHTSA predicted that automakers could reduce vehicle
weight as one part of the industry's strategy for meeting the proposed standards. EPA's
modeling projected (and still projects) that vehicle manufacturers will reduce the weight of
their vehicles by 4% on average between 2011 and 2016, with the average per-vehicle mass
reduction in absolute terms being greater for light-trucks than for passenger cars. For the
NPRM, this mass reduction estimate was generally smaller on both an absolute and a
percentage basis for smaller car than for larger vehicles. Specifically, we estimated an
average reduction of 2.3% (62 Ibs) for cars with a curb weight below 2,950 Ibs, 4.4% (154
Ibs) for cars with a curb weight above 2,950, 3.5% (119 Ibs) for trucks with a curb weight
below 3,850 Ibs, and 4.5% for trucks with a curb weight above 3,850 Ibs (215 Ibs). The
contained in a previous agency report. (The next citation in this document, Blincoe, L. and Shankar, U, January
2007.) The impacts were estimated on a year-by-year basis, but could be examined in a combined fashion. The
agency assumed that the safety trends will result in a reduction in the target population of fatalities from which
the weight impacts are derived. Using this method, we found a 12.6 percent reduction in fatality levels between
2007 and 2020. The estimates derived from applying Kahane's percentages to a baseline of 2007 fatalities were
thus multiplied by 0.874 to account for changes that the agency believes will take place in passenger car and
light truck safety between the 2007 baseline on-road fleet used for this particular analysis and year 2020.
KKK Blincoe, L. and Shankar, U, "The Impact of Safety Standards and Behavioral Trends on Motor Vehicle
Fatality Rates," DOT HS 810 777, January 2007. See Table 4 comparing 2020 to 2007 (37,906/43,363 = 12.6%
reduction.
7-136
-------
Environmental and Health Impacts
penetration and magnitude of these modeled changes are consistent with the public
announcements made by many manufacturers since early 2008 and are consistent with
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 March 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 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 platforms362
• Toyota: 30% weight reduction on 2015 Corolla and a 10% weight reduction on mid-
size vehicles by 2015363
• Nissan: 15% average weight reduction by 2015363
• Mazda: 100 kg (220 pound) weight reduction by 2011 and an additional 100 kg
weight reduction by 2016363'364
• Mercedes: 2-3 % weight reduction on recently introduced 2009 "BlueEFFICIENCY"
models365
The EPA believes that reducing vehicle mass of the magnitude we estimated for the proposal
and the final rule (up to 10% for certain vehicles by model year 2016)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 RIA. In
the NPRM, EPA described how 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. In their comments to the proposed rule, the American Iron and
Steel Institute (AISI) noted: "AISI has shown in its research with the Auto/Steel
Partnership and in programs supported by the U.S. Department of Energy, the use of
new Advanced High Strength Steel (AHSS) steel grades can enable the mass of
7-137
-------
Regulatory Impact Analysis
critical crash structures, such as front rails and bumper systems, to be reduced by
25%."
• 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 mass reduction that can be realized in the
powertrain is due to reduced powertrain output requirements. 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, mass
reduction 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
weight savings.366
The EPA stated in the NPRM that it believes that 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 mentioned that 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 goals. We requested 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 projected that automakers will not reduce vehicle footprint in response to the
proposed CO2 standards in our modeling analysis. NHTSA and EPA have taken two
measures to help ensure that this final rule does not provide an 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 rule is that
7-138
-------
Environmental and Health Impacts
the COi or fuel economy targets are based on the attribute of footprint (which is a surrogate
for vehicle size).LLL The second design feature is that the shape of the footprint curve (or
function) has been carefully designed with the intention that it neither encourages
manufacturers to increase, nor decrease the footprint of their fleet. Changes in relative safety
are related to shifts in the distribution of vehicles on the road. A policy that induces a
widening in the size distribution of vehicles on the road, could result in negative impacts on
safety. The primary mechanism in this rulemaking for mitigating the potential negative
effects on safety is the application of footprint-based standards, which create a disincentive
for manufacturers to produce smaller-footprint vehicles. This is because as footprint
decreases, the corresponding fuel economy/GHG emission target becomes more
stringent.MMM The shape of the footprint curves themselves have also been designed to be
approximately "footprint neutral" within the sloped portion of the functions - that is, to
neither encourage manufacturers to increase the footprint of their fleets, nor to decrease it.
Upsizing also is discouraged through a "cut-off at larger footprints. For both cars and light
trucks there is a "cut-off that affects vehicles smaller than 41 square feet. The agencies
recognize that for manufacturers who make small vehicles in this size range, this cut off
creates some incentive to downsize (i.e. further reduce the size and/or increase the production
of models currently smaller than 41 square feet) to make it easier to meet the target. The cut
off may also create some incentive for manufacturers who do not currently offer such models
to do so in the future. However, at the same time, the agencies believe that there is a limit to
the market for cars smaller than 41 square feet - most consumers likely have some minimum
expectation about interior volume, among other things. In addition, vehicles in this market
segment are the lowest price point for the light-duty automotive market, with a number of
models in the $10,000 to $15,000 range. In order to justify selling more vehicles in this
market in order to generate fuel economy or COi credits (that is, for this final rule to be the
incentive for selling more vehicles in this small car segment), a manufacturer would need to
add additional technology to the lowest price segment vehicles, which could be challenging.
Therefore, due to these two reasons (a likely limit in the market place for the smallest sized
cars and the potential consumer acceptance difficulty in adding the necessary technologies in
order to generate fuel economy and CO2 credits), the agencies believe that the incentive for
manufacturers to increase the sale of vehicles smaller than 41 square feet due to this
rulemaking, if present, is small. For further discussion on these aspects of the standards,
please see Section II.C above and Chapter 2 of the Joint TSD. However, EPA acknowledges
LLL 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. However, EPA does acknowledge that front and rear and side overhang
are not included in footprint and that there may be changes in this part of the vehicle in the future.
MMM we note, however, that vehicle footprint is not synonymous with vehicle size. Since the footprint is only
that portion of the vehicle between the front and rear axles, footprint-based standards do not discourage
downsizing the portions of a vehicle in front of the front axle and to the rear of the rear axle, or to other portions
of the vehicle outside the wheels. The crush space provided by those portions of a vehicle can make important
contributions to managing crash energy. At least one manufacturer has confidentially indicated plans to reduce
overhang as a way of reducing mass on some vehicles during the rulemaking time frame. Additionally, simply
because footprint-based standards create no incentive to downsize vehicles, does not mean that manufacturers
may not choose to do so if doing so makes it easier to meet the overall standard (as, for example, if the smaller
vehicles are so much lighter that they exceed their targets by much greater amounts).
7-139
-------
Regulatory Impact Analysis
some uncertainty regarding how consumer purchases will change in response to the vehicles
designed to meet the model years 2012-2016 standards. This could potentially affect the mix
of vehicles sole in the future, including the mass and footprint distribution, and thus result in a
different impact on safety than is discussed and presented in this final rule.
EPA also discussed safety in the context of NHTSA's analysis presented in Section IV
of the NPRM preamble, in which NHTSA presented an analysis of the proposed CAFE
standards based on the 2003 Kahane safety 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.367'36 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.
Based on this, NHTSA developed and presented a worst case estimate of the effect of weight
reductions on fatalities. The underlying data used for that analysis did not allow NHTSA to
analyze the specific effect 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 included vehicles that were either smaller and lighter or
larger and heavier. The numbers in the NHTSA analysis predicted 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 (as well as structural
integrity). EPA acknowledged that the safety analysis conducted by NHTSA and presented in
Section IV of the NPRM Preamble could be a worst case analysis for fatalities, but that the
actual effects on vehicle safety could (and would likely) be much less. However, EPA and
NHTSA were not able to quantify the lower-bound potential effects at that time.
Thus, the 2003 and earlier Kahane studies (as summarized in the NPRM) concerning
weight reductions and safety indicate that there is not a clear cut answer to the issue before the
agencies of potential safety impacts of the rule due to mass reduction which may occur in the
fleet absent size reductions. These studies draw upon historical vehicle data where mass
reduction were (for the most part) linked with size reductions, thus they occurred in ways
different from the way the agencies project weight reduction will occur for MY 2012-2016
vehicles. As such the pre-2010 Kahane studies may not be directly relevant to predicting or
quantifying the safety impact of the projected use of weight reduction technologies in this
rule. Several commenters to the proposed rule support this conclusion, including the
International Council on Clean Transportation (ICCT): "The results of the Kahane study
would be great for analyzing the safety impacts of a weight-based attribute system....
Kahane's methodology was simply not designed to assess the safety impacts of lightweight
materials.", and the National Association of Clean Air Agencies (NACAA): "significant
portions of NHTSA's safety analysis are based on out-of-date data".
In contrast to the pre-2010 Kahane studies, Dynamic Research Incorporated (DRI) has
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
7-140
-------
Environmental and Health Impacts
studies,369'370 >NNN DRI presented results that indicated that vehicle weight reduction tended to
decrease fatalities, but vehicle wheelbase and track reduction tended to increase fatalities.
In the RIA for the proposed rule, EPA attempted to summarize DRI's results in four
major points. These points are re-summarized below to reflect comments from DRI as to
present this information more accurately:000
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 and/or increases in wheel base and track.
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 a potentially
lower center of gravity and lower loads on the roof structures.
The data used for the DRI analysis was similar to that used in 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 1998 vehicles. DRI also used a logistic 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. NHTSA stated
that 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.ppp 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. 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.
NNN One of these studies was published as a Society of Automotive Engineers Technical Paper and received peer
review through that body
000 DRI stated: "The DRI work focused on the effects of vehicle size and weight (i.e. curb weight, wheelbase,
and track) on vehicle crash avoidance, crashworthiness, and compatibility, based on accident and fatality data.
There were numerous conclusions in addition to those listed, including the benefits of both increased size and
weight."
ppp 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.
7-141
-------
Regulatory Impact Analysis
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 (along with many commenters) recognized 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 in the 2003 Kahane study
above.
The DRI and Kahane studies also differed 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
indicated 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 offered 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.
In the NPRM, NHTSA rejected many elements of the DRI analysis, and did not rely
on it for its evaluation of safety impact changes from the proposed CAFE standards. See
Section IV.G.6 of the Notice of Proposed Rulemaking, as well as NHTSA March 2009 Final
Rulemaking for MY2011 CAFE standards (74 FR at 14402-05).
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 noted that if DRI's results were to be applied using the curb
weight reductions predicted by the OMEGA model, an overall reduction in fatalities would be
predicted. There were many commenters who supported this notion. Further discussion of
these comments is included below.
7.6.2 What Public Comments did EPA Receive in Regard to its Safety Discussion
and what is EPA's Response?
EPA requested and received public comments from several sources regarding the
NPRM Preamble positions noted above. Many of the comments received were in support of
EPA's safety assessment, such as those from Dynamics Research Institute (DRI),
International Council on Clean Transportation (ICCT), Public Citizen, Union of Concerned
Scientists, California Air Resources Board (CA-ARB) and the National Association of Clean
Air Agencies (NACAA).
In a technical comment, DRI's agreed with EPA that directional control and therefore
crash avoidance improves with a reduction in curb weight. DRI went further and commented
7-142
-------
Environmental and Health Impacts
that control also improved with an increase in wheel base and track width., The latter point
acknowledges that a corresponding potential decrease in wheel base or track width could
result in an increase in fatalities. DRI offered results from the application of the "quasi-steady
vehicle directional equations of motion". These results show "that passenger cars with shorter
wheelbases tend to have smaller characteristic speeds, resulting in higher [yaw rate] natural
frequency and less damping, based on analysis of quasi-steady vehicle equations of motion".
EPA concurs with DRI's comments regarding their analysis. This supports the
argument made by EPA in the NPRM that, vehicles with reduced mass are better able to avoid
accidents that cause fatalities due to increased vehicle maneuverability that better matches a
driver's intended steering inputs.QQQ EPA analyzed Consumer Reports Double Lane Change
DataRRR to determine how curb weight affects emergency handling. Consumer Reports
provided EPA with a summary of their double lane change tests for vehicles from MY 2003
through MY 2010 along with the vehicles' curb weight.888 By plotting the vehicle double
lane change speed with respect to curb weight, EPA is able to show how the speed at which a
double lane change can be safely executed goes up as curb weight decreases, as shown in
Figure 7-29 below. As the weight of all vehicles is decreased, their maneuverability and
handling will increase, thus giving the driver greater control and increased capability to avoid
accidents. This relationship held true for both vehicles equipped with and without Electronic
Stability Control (ESC), although the same data show that capability is further enhanced with
the addition of ESC. At any given curb weight, the addition of ESC resulted in a higher
average double lane change speed than those vehicles without stability control. Under all
stability control events, the ESC system will be working against the momentum of the
vehicle. Reductions in mass will generally increase ESC effectiveness in terms of the ability
of a vehicle to enter a double lane change maneuver at a higher speed. However, NHTSA's
review of historical EARS data for MY 1991-1999 cars has found that smaller cars are
responsible for a higher number of crashes than other vehicle types.
QQQ FMVSS126 - "Oversteering and understeering are typically cases of loss-of-control where vehicles move in
a direction different from the driver's intended direction".
RRR The Consumer Reports double lane change maneuver is referred to as "Emergency Handling" within
Consumer Reports vehicle assessments. The maneuver consists of a set of traffic cones arranged in such a
pattern as to force the vehicle into a left lane, and then a return to the right lane over a controlled distance. The
double lane change speed is based on the driver's entrance speed at the beginning of the course. The highest
entrance speed with which the driver is able to negotiate the course without disturbing the cones is deemed the
"Emergency Handling" metric.
sss The data used for Figure 7.6.2 is available in the memorandum "Vehicle Double Lane Change Data Provided
by Consumer Reports", which has been placed in the EPA docket for this final rule.
7-143
-------
Regulatory Impact Analysis
Vehicles equipped with ESC
yv0.0026x + 62.114
R2 = 0.5282
o o o
X O XOD
O »O O O OO
OO O OOD OO
O CD
o
X X CT»Offl> 000
x ox xx o»ac
X O OK K*W)0000
ODXXK
Vehicles without ESC
y = -0.0025x + 59.921
R2 = 0.4085
X Non-ESC
o ESC Vehicles
Linear (ESC Vehicles)
Linear (Non-ESC)
1000
2000
3000 4000 5000
Vehicle Curb Weight (Ibs)
6000
7000
Figure 7-29 Double Lane Change Speed vs. Vehicle Curb Weight
The International Council on Clean Transportation (ICCT) endorsed the DRI
comments (described in the previous section) in their comments to the proposal. ICCT went a
step further and applied the DRI results to calculate a potential benefit from reducing curb
weight while maintaining footprint. Accounting for the downward trend in annual vehicle
fatalities, and the actual weight reduction anticipated by agencies' modeling results, ICCT
determined that there would be a reduction of 599 fatalities and 354 fatalities for a 100 Ib
weight reduction in cars and trucks respectively.
In contrast, there were comments received from the American Iron and Steel Institute
(AISI) and the Competitive Enterprise Institute (CEI) that contended EPA's conclusions.
Both AISI and CEI noted in their comments that historically heavier vehicles have
demonstrated a safety advantage over lighter vehicles. AISI presented the results from
Desapriya that "sedans are two times more likely to be injured than drivers or passengers in
larger pickup trucks and SUV's", and CEI endorsed the results from the 2003 Kahane study in
support of their position.
AISI also noted that: "AISI has shown in its research with the Auto/Steel Partnership
and in programs supported by the U.S. Department of Energy, the use of new (Advanced High
Stength Steel) AHSS steel grades can enable mass of critical crash structures, such as front
rails and bumper systems, to be reduced by 25 percent. Such vehicle structures with reduced
mass can perform as well as their heavier counterparts in standard NHTSA frontal or IIHS
offset instrumented crash tests." This is exactly the type of material substitution that EPA
7-144
-------
Environmental and Health Impacts
noted in the NPRM. In this case a vehicle can be made lighter without a concomitant
decrease in crush and crumple zones.
CEI noted the historical studies that analyzed the relationships between CAFE
standards and vehicle fatalities, specifically, R.W. Crandall and J.D. Graham, authors of "The
Effect of Fuel Economy Standards on Automobile Safety" and the 2003 Kahane report.
According to CEI, the Crandall and Graham analysis reported that CAFE regulations had a
downsizing effect of 500 pounds per car in 1989. The applicability of the 1989 analyses is
questionable in the context of this rule and its associated effectiveness from MY 2012 through
2016. Vehicle design capability, including computer aided engineering (CAE) has changed
substantially in the 20 years since this report was completed. Increased stringency of FMVSS
standards combined with the improved strength of materials available and used in automobile
manufacturing has allowed auto manufacturers to improve impact performance without a
commensurate increase in vehicle weight. In addition, the 500 pound reduction analyzed by
Crandall and Graham is substantially greater than the estimated weight reduction associated
with this final rule. Regarding the 2003 Kahane report, NHTSA has recognized some of the
limitations of the 2003 analysis in the context of this rulemaking and presents a summary and
review of the 2010 Kahane report in the preamble and within NHTSA's FRIA.
CEI further comments that "new technologies and attribute-based regulation will not
eliminate the safety tradeoff. CEI states that "some proponents of higher CAFE standards,
and of CO2 emission limits, claim that new technologies can eliminate these lethal effects.
This claim is simply false, even if such technologies do not themselves involve downsizing.
Consider a hi-tech prototype car capable of meeting either a higher CAFE standard, or a
stringent CO2 emissions standard. Imagine that you then increase this car's size and weight
by adding several cubic feet of trunk space and occupant space. The result would be an even
safer car." EPA recognizes that more recent vehicles have more safety features than 1990s
vehicles, which are likely to make them safer overall. To account for this, NHTSA did adjust
the results of both its NPRM and final rule analysis to include known safety improvements,
like ESC and increases in seat belt use, that have occurred since MYs 1991-1999.TTT
However, simply because newer vehicles have more safety countermeasures, does not mean
that the weight/safety relationship necessarily changes. More likely, it would change the
target population (the number of fatalities) to which one would apply the weight/safety
relationship. Thus, EPA acknowledges that while mass reduction can be done in a safety
neutral manner, some mass reduction techniques for both passenger cars and light trucks can
make them less safe in certain crashes as discussed in NHTSA's FRIA.UUU
TTT See Chapter IX of the NHTSA FRIA for details on this adjustment.
uuu If one has a vehicle (vehicle A), and both reduces the vehicle's mass and adds new safety equipment to it,
thus creating a variant (vehicle Aj), the variant might conceivably have a level of overall safety for its occupants
is equal to that of the original vehicle (vehicle A). However, vehicle Aj might not be as safe as second variant
(vehicle A2) of vehicle A, one that is produced by adding to vehicle A the same new safety equipment added to
the first variant, but this time without any mass reduction.
7-145
-------
Regulatory Impact Analysis
Additional comments received from the AISI challenge the agencies' conclusions with
regard to vehicle size. While they agree that "the historical relationship between vehicle size,
weight, and collision severity may be influenced by design and structural improvements",
they assert that "the aggressive schedule for implementing the proposed rule assures that
carmakers will be manufacturing smaller, lighter vehicles in order to comply." Porsche and
IIHS had similar comments implying that footprint standards will increase the risk that
manufacturers will make vehicles smaller. In response, EPA does not feel that the schedule is
overly aggressive. The standards and their phase-in period is feasible and economically
practical as described in section III.D of the preamble. That same discussion, as well as the
supporting analysis in the joint TSD, demonstrates a clear compliance path to meeting the
standards based on wider penetration of existing technologies, at reasonable and affordable
cost. In addition, the attribute-based approach, along with the shape of footprint curve has
been developed with the objective of minimizing the incentive to downsize (since, among
other things, downsizing simply creates a more stringent regulatory target corresponding to
the downsized footprint) nor to upsize (as that could incur significant redesign costs and also
would likely increase the mass of the vehicle which could off-set any benefit in upsizing
from a stringency perspective as a higher mass vehicle will produce more CO2 emissions).
EPA consequently does not accept the commenter's assertion. However, as discussed in
Sections III.H.l and IV.G.6 of the preamble, the agencies acknowledge some uncertainty
regarding how consumer purchases will change in response to the vehicles designed to meet
the MYs 2012-2016 standards. This could potentially affect the mix of vehicles sold in the
future, including the mass and footprint distribution.
7.6.3 NHTSA's 2010 Study of Accident Fatalities by Vehicle Size and Weight
In response to comments received and in an effort to follow through on some of its
NRPM pledges for consideration in the future, NHTSA has significantly revised its 2003
study on the relationship between vehicle mass and fatalities in the context of this 2012-2016
rulemaking. A copy of this new report, "Relationships Between Fatality Risk, Mass, and
Footprint in Model Year 1991-1999 and Other Passenger Cars and LTVs", Charles J. Kahane,
NCSA, NHTSA, March 2010, has been placed in the docket for this rulemaking, hereafter
referred to as the 2010 Kahane report.v In the new 2010 Kahane report, based on the
original MY 1991-1999 vehicle data set, NHTSA addresses several criticisms, specifically
including 2-door vehicles in its passenger car analysis, attempting to separate weight and
footprint as independent variables and their associated contributions to vehicle fatalities, and
comparing their results to those from DRI. We note that this analysis looks specifically at
impacts on fatalities. EPA and NHTSA have not analyzed the impact of mass reduction
predicted from this final rule on non-fatal accidents. We also note that the results of the 2010
Kahane analysis, as applied to the mass reductions predicted from this rulemaking, must be
viewed in the overall context of our projection of the 2012-2016 new vehicle fleet distribution
(See Chapter 1 of the Joint TSD for a discussion of how EPA and NHTSA have developed the
NHTSA intends for this 2010 Kahane report to undergo a peer review in accordance with OMB guidance
for peer review, and the results of this peer review and any subsequent revisions to the report will be made
available to the public upon completion.
7-146
-------
Environmental and Health Impacts
projected future new vehicle fleet). EPA acknowledges some uncertainty regarding how
consumer purchases may change in response to the vehicles designed to meet the MYs 2012-
2016 standards. This could potentially affect the mix of vehicles sold in this time frame,
including the mass and footprint distribution - which would impact the projection of fatalities
from the 2010 Kahane analysis as applied to this final rule.
The 2010 Kahane report presents three sets of results each with an estimate of
fatalities based on the weight reductions projected in the feasibility analysis for the final rule.
Each of these results presents a significant departure from the NPRM. The first set of results
is a straight regression of result for passenger cars and LTVs. The passenger car data now
includes 2-door vehicles, although a decision was made to continue to exclude muscle cars.
The second and third sets of results, termed "upper bound scenario" and "lower bound
scenario" respectively, are the result of expert opinion and judgment by NHTS A as to how
mass may be reduced and the potential effects, both primary for the driver and occupants of a
vehicle, and secondary or societal effects. NHTSA was able to perform these analyses with
both mass and footprint treated as independent variables in the 2010 study. The unmodified
straight regression results now reflect an expected reduction of 301 fatalities for the life of
MY 2012-2016 vehicles, down from the absolute worst-case 493 in the CAFE NPRM. The
new "upper bound scenario" estimates that there will be 22 additional fatalities as the result of
the CAFE rule, and the "lower bound scenario" an 80 fatality decrease***. These results are
consistent with EPA's NPRM claim that fatalities as a result of the rule could be close to zero.
However, NHTSA states in the new 2010 Kahane analysis that the potential fatality increases
associated with mass reduction in the passenger cars would be to a large extent offset by the
benefits of mass reduction in the heavier LTVs. As was stated in the NPRM, EPA continues
to believe that weight can be reduced from passenger cars safely through smart design, and
other methods which can be used in the model years 2012 to 2016 time frame. This is based
on a number of studies in the literature including those from DRI, Wenzel, Ross and
Robertson referenced above and elsewhere in this final rule.
Furthermore, in an effort to address public comments that promoted the DRI results as
a legitimate alternative to NHTS A's analysis, the 2010 Kahane report presents several
conclusions. NHTSA first focused on the issue of "near multicollinearity" of the data. This
statistical characteristic can result in increased uncertainty of regression coefficients, which
according to sources cited by NHTSA, can result in the "wrong sign or implausible
magnitude". There are statistical tests that can be run on data used in regression models to
determine the level of multicollinearity between independent variables. NHTSA performed
tests on their data set and determined that mass and footprint did exhibit "near
multicollinearity". Subsequently, NHTSA applied a 2-step regression, in accordance with
DRI's methodology, to its own data and the results showed a corresponding decrease in
fatalities for a 100 pound mass reduction. NHTSA did not accept these results for many
The "upper-estimate scenario" and "lower-estimate scenario" are based on NHTS A's judgment as a vehicle
safety agency, and are not meant to convey a sense of confidence in the precision of the results, but more to
convey a sense of bounding for potential safety effects.
7-147
-------
Regulatory Impact Analysis
reasons, as outlined in the 2010 Kahane report, however their underlying conclusion was that
the 2-step regression applied by DRI exacerbated the effects of "near multicollinearity".
The EPA has found one additional peer-reviewed study of historical PARS data.
Robertson 2006371 analyzed somewhat newer model year passenger cars and light-trucks
(1999-2002) and used a logistic regression approach similar to the 2003 Kahane and 2005
DRI studies, including consideration of driver gender and age. We note that this peer
reviewed paper was published as a commentary in response to previous work published by the
author, and this commentary does include new original work. The reference list contained in
this 2006 Robertson paper includes the references to the previous work by the author The
study found multicollinearity to be a problem for regressions, including vehicle curb-weight
and wheelbase, but found turning radius could be substituted as an indicator of vehicle size
without introducing significant multicollinearity into the regression analysis. Robertson's
analysis tested the hypothesis of reducing vehicle curb-weight to the minimum achievable by
the 1999-2002 model year population vs. vehicle size. The regression results showed a
societal benefit of a 28% reduction in fatalities for minimization of mass. Mass minimization
vs a vehicle size metric is not the same as removing a fixed percent or fixed quantity of
vehicle mass and thus the results are not directly comparable to either the 2010 Kahane or
2005 DRI results, but they are directionally consistent with both studies. The analysis showed
similar trends to the 2010 Kahane report with increased risk of fatality to drivers of lighter
vehicles, which was more than offset by the reduction in risk of fatality to other drivers,
similar to Kahane's 2010 results with respect to light trucks and to DRI's results for both
light-trucks and passenger cars.
The EPA believe that while NHTSA's new 2010 analysis significantly adds to the
literature and understanding of the effects of mass reduction on safety, that there still are
many opportunities for further study.
7.6.4 Suggested Next Steps To Increase Our Understanding of the Effects of
Vehicle Size and Weight on Fatalities
NHTSA and EPA believe that it is important for the agencies to conduct further study
and research into the interaction of mass, size and safety to assist future rulemakings. The
agencies intend to begin working collaboratively and to explore with DOE, CARB, and
perhaps other stakeholders an interagency/ intergovernmental working group to evaluate all
aspects of mass, size and safety. It would also be the goal of this team to coordinate
government supported studies and independent research, to the extent possible, to help ensure
the work is complementary to previous and ongoing research and to guide further research in
this area. DOE's EERE office has long funded extensive research into component advanced
vehicle materials and vehicle mass reduction. Other agencies may have additional expertise
that will be helpful in establishing a coordinated work plan. The agencies are interested in
looking at the weight-safety relationship in a more holistic (complete vehicle) way, and
thanks to this CAFE rulemaking NHTSA has begun to bring together parts of the agency—
crashworthiness, and crash avoidance rulemaking offices and the agency's Research &
Development office—in an interdisciplinary way to better leverage the expertise of the
agency. Extending this effort to other agencies will help to ensure that all aspects of the
weight-safety relationship are considered completely and carefully with our future research.
7-148
-------
Environmental and Health Impacts
The agencies also intend to carefully consider comments received in response to the NPRM in
developing plans for future studies and research and to solicit input from stakeholders.
The agencies also plan to watch for safety effects as the U.S. light-duty vehicle fleet
evolves in response both to the CAFE/GHG standards and to consumer preferences over the
next several years. Additionally, as new and advanced materials and component smart
designs are developed and commercialized, and as manufacturers implement them in more
vehicles, it will be useful for the agencies to learn more about them and to try to track these
vehicles in the fleet to understand the relationship between vehicle design and injury/fatality
data. Specifically, the agencies intend to follow up with study and research of the following:
First, NHTSA is in the process of contracting with an independent institution to
review the statistical methods that NHTSA and DRI have used to analyze historical data
related to mass, size and safety, and to provide recommendation on whether the existing
methods or other methods should be used for future statistical analysis of historical data. This
study will include an consideration of potential multicollinearity in the historical data and how
best to address it in a regression analysis. This study is being initiated because, in response to
the NPRM, NHTSA received a number of comments related to the methodology NHTSA
used for the NPRM to determine the relationship between mass and safety, as discussed in
detail above.
Second, NHTSA and EPA, in consultation with DOE, intend to begin updating the
MYs 1991-1999 database on which the safety analyses in the NPRM and final rule are based
with newer vehicle data in the next several months. This task will take at least a year to
complete. This study is being initiated in response to the NPRM comments related to the use
of data from MYs 1991-1999 in the NHTSA analysis, as discussed in the section II.G of the
preamble.
Third, in order to assess if the design of recent model year vehicles that incorporate
various mass reduction methods affect the relationships among vehicle mass, size and safety,
NHTSA and EPA intend to conduct collaborative statistical analysis, beginning in the next
several months. The agencies intend to work with DOE to identify vehicles that are using
material substitution and smart design. After these vehicles are identified, the agencies intend
to assess if there are sufficient data for statistical analysis. If there are sufficient data,
statistical analysis would be conducted to compare the relationship among mass, size and
safety of these smart design vehicles to vehicles of similar size and mass with more traditional
designs. This study is being initiated because, in response to the NPRM, NHTSA received
comments related to the use of data from MYs 1991-1999 in the NHTSA analysis that did not
include new designs that might change the relationship among mass, size and safety, as
discussed in detail above.
NHTSA may initiate a two-year study of the safety of the fleet through an analysis of
the trends in structural stiffness and whether any trends identified impact occupant injury
response in crashes. Vehicle manufacturers may employ stiffer light weight materials to limit
occupant compartment intrusion while controlling for mass that may expose the occupants to
higher accelerations resulting in a greater chance of injury in real-world crashes. This study
7-149
-------
Regulatory Impact Analysis
would provide information that would increase the understanding of the effects on safety of
newer vehicle designs.
In addition, EPA and NHTSA, possibly in collaboration with DOE, may conduct a
longer-term computer modeling-based design and analysis study to help determine the
maximum potential for mass reduction in the MYs 2017-2021 timeframe, through direct
material substitution and smart design while meeting safety regulations and guidelines, and
maintaining vehicle size and functionality. This study may build upon prior research
completed on vehicle mass reduction. This study would further explore the comprehensive
vehicle effects, including dissimilar material joining technologies, manufacturer feasibility of
both supplier and OEM, tooling costs, and crash simulation and perhaps eventual crash
testing.
7-150
-------
Environmental and Health Impacts
References
References can be found in the EPA DOCKET (EPA-HQ-OAR-2009-0472) or are
publicly available.
1 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Docket
EPA-HQ-OAR-2009-0472-11295.
2 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section
2.3.1.1. Docket EPA-HQ-OAR-2009-0472-11295.
3 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section
2.3.1.2. Docket EPA-HQ-OAR-2009-0472-11295.
4 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section
2.3.4. Docket EPA-HQ-OAR-2009-0472-11295.
5 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Table 2-6.
Docket EPA-HQ-OAR-2009-0472-11295.
6 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section
2.3.5.1. Docket EPA-HQ-OAR-2009-0472-11295.
7 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Table 2-6.
Docket EPA-HQ-OAR-2009-0472-11295.
o
U.S. EPA. (2006 j. 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/. Docket EPA-HQ-
OAR-2009-0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-
0101.
9 U.S. EPA. (2007 j. 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/. Docket EPA-HQ-OAR-2009-0472-0106.
10 National Research Council (NRC), 2008. Estimating Mortality Risk Reduction and
Economic Benefits from Controlling Ozone Air Pollution. The National Academies Press:
Washington, D.C. Docket EPA-HQ-OAR-2009-0472-0322.
11 Bates, D.V., Baker-Anderson, M., Sizto, R. (1990). Asthma attack periodicity: a study of
hospital emergency visits in Vancouver. Environ. Res., 57,51-70. Docket EPA-HQ-OAR-
2009-0472-0264.
7-151
-------
Regulatory Impact Analysis
12Thurston, 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.
Docket EPA-HQ-OAR-2009-0472-0265.
13Thurston, 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. Docket EPA-HQ-OAR-2009-0472-0266.
14Lipfert, F.W., Hammerstrom, T. (1992). Temporal patterns in air pollution and hospital
admissions. Environ. Res., 59,374-399. Docket EPA-HQ-OAR-2009-0472-0267.
15 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. Docket EPA-HQ-OAR-2009-0472-0268.
16 U.S. EPA. (2006 j. 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/. Docket EPA-HQ-
OAR-2009-0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-
0101.
17 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/. Docket EPA-HQ-
OAR-2009-0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-
0101.
18 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. Cell Mol. Biol., 4, 72-81.
19 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. Docket EPA-HQ-OAR-2009-0472-
0269.
20 Koren, 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. Docket EPA-HQ-OAR-2009-0472-0323.
21 Schelegle, E.S., Siefkin, A.D., McDonald, RJ. (1991). Time course of ozone-induced
neutrophilia in normal humans. Am. Rev. Respir. Dis., 743,1353-1358. Docket EPA-HQ-
OAR-2009-0472-0324.
22 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. Docket EPA-HQ-OAR-2009-0472-0099, EPA-HQ-OAR-
2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
7-152
-------
Environmental and Health Impacts
23Hodgkin, I.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. Docket
EPA-HQ-OAR-2009-0472-0270.
24Euler, G.L., Abbey, D.E., Hodgkin, I.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. Docket EPA-HQ-OAR-2009-0472-0360.
25 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. Docket EPA-HQ-
OAR-2009-0472-0361.
26 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/. Docket EPA-HQ-OAR-2009-0472-0106.
27 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/. Docket EPA-HQ-
OAR-2009-0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-
0101.
28 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/. Docket EPA-HQ-
OAR-2009-0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-
0101.
29 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. Docket EPA-HQ-OAR-2009-0472-0362.
30Higgins, 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. Docket EPA-HQ-OAR-2009-0472-0332.
31 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. Docket EPA-HQ-OAR-2009-0472-0363.
32 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. Docket
EPA-HQ-OAR-2009-0472-0364.
33 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
7-153
-------
Regulatory Impact Analysis
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). Docket EPA-HQ-OAR-2009-0472-0383.
34Spektor, 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. Docket EPA-HQ-OAR-2009-0472-0380.
35 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., 137, 313-320. Docket EPA-HQ-OAR-2009-0472-
0333.
36 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/. Docket EPA-HQ-
OAR-2009-0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-
0101.
37Hazucha, 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.
Docket EPA-HQ-OAR-2009-0472-0334.
38Horstman, 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.
39Horstman, 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. Docket EPA-HQ-OAR-2009-0472-0384.
40 U.S. EPA (2008). Integrated Science Assessment (ISA) for Sulfur Oxides - Health Criteria
(Final Report). EPA/600/R-08/047F. Washington, DC,: U.S.EPA. Retrieved on March 19,
2009 from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=198843. Docket EPA-HQ-
OAR-2009-0472-0335.
41 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. Docket EPA-HQ-
OAR-2009-0472-0350.
42 U.S. EPA, 2010. Integrated Science Assessment for Carbon Monoxide (Final Report). U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R-09/019F, 2010.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686. Docket EPA-HQ-OAR-2009-
0472-0335.
43 U. S. EPA. (2009) 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata2002/risksum.html Docket EPA-HQ-OAR-2009-0472-
11322.
7-154
-------
Environmental and Health Impacts
44 U.S. EPA (2007) Control of Hazardous Air Pollutants from Mobile Sources. 72 FR 8428;
February 26, 2007. Docket EPA-HQ-OAR-2009-0472-0271.
45 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. Docket
EPA-HQ-OAR-2009-0472-0391.
46 U.S. EPA (2009) National-Scale Air Toxics Assessment for 2002. This material is
available electronically at http://www.epa.gov/ttn/atw/nata2002/risksum.html. Docket EPA-
HQ-OAR-2009-0472-11322.
47 U.S. EPA (2009) National-Scale Air Toxics Assessment for 2002.
http://www.epa.gov/ttn/atw/nata2002. Docket EPA-HQ-OAR-2009-0472-11321.
48 U.S. EPA. 2000. Integrated Risk Information System File for Benzene. This material is
available electronically at: http://www.epa.gov/iris/subst/0276.htm. Docket EPA-HQ-OAR-
2009-0472-1659.
49 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. Docket EPA-HQ-OAR-2009-0472-0366.
50 Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry, V.A. (1992) Synergistic action of
the benzene metabolite hydroquinone on myelopoietic stimulating activity of
granulocyte/macrophage colony-stimulating factor in vitro, Proc. Natl. Acad. Sci. 89:3691-
3695. Docket EPA-HQ-OAR-2009-0472-0370.
51 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. EPA-HQ-OAR-2009-
0472-0367. Docket EPA-HQ-OAR-2009-0472-0367.
52 U.S. Department of Health and Human Services National Toxicology Program llth Report
on Carcinogens available at: http://ntp.niehs.nih.gov/go/16183.
53Aksoy, M. (1989). Hematotoxicity and carcinogenicity of benzene. Environ. Health
Perspect. 82: 193-197. EPA-HQ-OAR-2009-0472-0368
54 Goldstein, B.D. (1988). Benzene toxicity. Occupational medicine. State of the Art
Reviews. 3: 541-554. Docket EPA-HQ-OAR-2009-0472-0325.
55 Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E. Marti, Y.Z. Wang, M. Linet,
L.Q. Xi, W. Lu, M.T. Smith, N. Titenko-Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B.
Hayes (1996) Hematotoxicity among Chinese workers heavily exposed to benzene. Am. J.
Ind. Med. 29: 236-246. Docket EPA-HQ-OAR-2009-0472-0326.
56 U.S. EPA 2002 Toxicological Review of Benzene (Noncancer Effects). Environmental
Protection Agency, Integrated Risk Information System (IRIS), Research and Development,
National Center for Environmental Assessment, Washington DC. This material is available
electronically at http://www.epa.gov/iris/subst/0276.htm. Docket EPA-HQ-OAR-2009-0472-
0327.
7-155
-------
Regulatory Impact Analysis
57 Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, B.; Melikian, A.; Eastmond, D.;
Rappaport, S.; Li, H.; Rupa, D.; Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.;
Winnik, M.; Kwok, E.; Li, Y.; Mu, R.; Xu, B.; Zhang, X.; Li, K. (2003). HEI Report 115,
Validation & Evaluation of Biomarkers in Workers Exposed to Benzene in China. Docket
EPA-HQ-OAR-2009-0472-0328.
58 Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et al. (2002). Hematological changes
among Chinese workers with a broad range of benzene exposures. Am. J. Industr. Med. 42:
275-285. Docket EPA-HQ-OAR-2009-0472-0329.
59 Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004). Hematotoxically in Workers
Exposed to Low Levels of Benzene. Science 306: 1774-1776. Docket EPA-HQ-OAR-2009-
0472-0330.
60 Turtletaub, K.W. and Mani, C. (2003). Benzene metabolism in rodents at doses relevant to
human exposure from Urban Air. Research Reports Health Effect Inst. Report No. 113.
Docket EPA-HQ-OAR-2009-0472-0385.
61 U.S. EPA. 2002. Health Assessment of 1,3-Butadiene. Office of Research and
Development, National Center for Environmental Assessment, Washington Office,
Washington, DC. Report No. EPA600-P-98-001F. This document is available electronically
at http://www.epa.gov/iris/supdocs/buta-sup.pdf. Docket EPA-HQ-OAR-2009-0472-0386.
62 U.S. EPA. 2002 "Full IRIS Summary for 1,3-butadiene (CASRN 106-99-0)" Environmental
Protection Agency, Integrated Risk Information System (IRIS), Research and Development,
National Center for Environmental Assessment, Washington, DC
http://www.epa.gov/iris/subst/0139.htm. Docket EPA-HQ-OAR-2009-0472-1660.
63 International Agency for Research on Cancer (IARC) (1999) Monographs on the evaluation
of carcinogenic risk of chemicals to humans, Volume 71, Re-evaluation of some organic
chemicals, hydrazine and hydrogen peroxide and Volume 97 (in preparation), World Health
Organization, Lyon, France. Docket EPA-HQ-OAR-2009-0472-0387.
64 U.S. Department of Health and Human Services National Toxicology Program llth Report
on Carcinogens available at: http://ntp.niehs.nih.gov/go/16183.
65Bevan, C.; Stadler, J.C.; Elliot, G.S.; et al. (1996) Subchronic toxicity of 4-
vinylcyclohexene in rats and mice by inhalation. Fundam. Appl. Toxicol. 32:1-10. Docket
EPA-HQ-OAR-2009-0472-0388.
66 U.S. EPA. 1987. Assessment of Health Risks to Garment Workers and Certain Home
Residents from Exposure to Formaldehyde, Office of Pesticides and Toxic Substances, April
1987. Docket EPA-HQ-OAR-2009-0472-0389.
67Hauptmann, M..; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2003. Mortality from
lymphohematopoetic malignancies among workers in formaldehyde industries. Journal of the
National Cancer Institute 95: 1615-1623. Docket EPA-HQ-OAR-2009-0472-0336.
68Hauptmann, M..; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2004. Mortality from
solid cancers among workers in formaldehyde industries. American Journal of Epidemiology
159: 1117-1130. Docket EPA-HQ-OAR-2009-0472-0337.
7-156
-------
Environmental and Health Impacts
69 Beane Freeman, L. E.; Blair, A.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Hoover, R. N.;
Hauptmann, M. 2009. Mortality from lymphohematopoietic malignancies among workers in
formaldehyde industries: The National Cancer Institute cohort. J. National Cancer Inst. 101:
751-761. Docket EPA-HQ-OAR-2009-0472-0338.
70 Pinkerton, L. E. 2004. Mortality among a cohort of garment workers exposed to
formaldehyde: an update. Occup. Environ. Med. 61: 193-200. Docket EPA-HQ-OAR-2009-
0472-0339.
71 Coggon, D, EC Harris, J Poole, KT Palmer. 2003. Extended follow-up of a cohort of British
chemical workers exposed to formaldehyde. J National Cancer Inst. 95:1608-1615. Docket
EPA-HQ-OAR-2009-0472-0340.
72Conolly, 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. Docket EPA-HQ-OAR-2009-0472-0341.
73 Conolly, 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. Docket EPA-HQ-OAR-2009-0472-0342.
74 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. Docket EPA-HQ-OAR-2009-0472-
0429.
75 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. Docket EPA-HQ-OAR-2009-0472-0369.
76 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. Docket EPA-HQ-OAR-2009-0472-0371.
77 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. Docket EPA-HQ-OAR-2009-0472-0444.
78 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.
Docket EPA-HQ-OAR-2009-0472-0447.
79 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.
Docket EPA-HQ-OAR-2009-0472-0447.
80 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. Docket EPA-HQ-OAR-2009-0472-0444.
7-157
-------
Regulatory Impact Analysis
81 International Agency for Research on Cancer (2006) Formaldehyde, 2-Butoxyethanol and
l-tert-Butoxypropan-2-ol. Monographs Volume 88. World Health Organization, Lyon,
France. Docket EPA-HQ-OAR-2009-0472-1164.
82 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 1 l.html. Docket EPA-HQ-OAR-
2009-0472-1191.
83 WHO (2002) Concise International Chemical Assessment Document 40: Formaldehyde.
Published under the joint sponsorship of the United Nations Environment Programme, the
International Labour Organization, and the World Health Organization, and produced within
the framework of the Inter-Organization Programme for the Sound Management of
Chemicals. Geneva. Docket EPA-HQ-OAR-2009-0472-1199.
84 U.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. Docket EPA-
HQ-OAR-2009-0472-0390.
85 U.S. Department of Health and Human Services National Toxicology Program llth Report
on Carcinogens available at: http://ntp.niehs.nih.gov/go/16183.
86 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.
87 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. Docket EPA-HQ-OAR-
2009-0472-0390.
88 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. Docket EPA-
HQ-OAR-2009-0472-0391.
89 Appleman, L.M., R.A. Woutersen, and VJ. Feron. (1982). Inhalation toxicity of
acetaldehyde in rats. I. Acute and subacute studies. Toxicology. 23: 293-297. Docket EPA-
HQ-OAR-2009-0472-0392.
90 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. Docket EPA-HQ-OAR-2009-0472-0408.
91 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. Docket EPA-HQ-OAR-2009-0472-0391.
92 International Agency for Research on Cancer (IARC). 1995. Monographs on the evaluation
of carcinogenic risk of chemicals to humans, Volume 63, Dry cleaning, some chlorinated
solvents and other industrial chemicals, World Health Organization, Lyon, France. Docket
EPA-HQ-OAR-2009-0472-0393.
7-158
-------
Environmental and Health Impacts
93 Sim VM, Pattle RE. Effect of possible smog irritants on human subjects JAMA165: 1980-
2010, 1957. Docket EPA-HQ-OAR-2009-0472-0395.
94 U.S. EPA (U.S. Environmental Protection Agency). (2003) Toxicological review of
acrolein in support of summary information on Integrated Risk Information System (IRIS)
National Center for Environmental Assessment, Washington, DC. EPA/635/R-03/003.
Available online at: http://www.epa.gov/ncea/iris.
95 Weber-Tschopp, A; Fischer, T; Gierer, R; et al. (1977) Experimentelle reizwirkungen von
Acrolein auf den Menschen. Int Arch Occup Environ Hlth 40(2): 117-130. In German. Docket
EPA-HQ-OAR-2009-0472-0394.
96 Integrated Risk Information System File of Acrolein. Office of Research and Development,
National Center for Environmental Assessment, Washington, DC. This material is available
at http://www.epa.gov/iris/subst/0364.htm. Docket EPA-HQ-OAR-2009-0472-0391.
97 U.S. EPA (U.S. Environmental Protection Agency). (2003) Toxicological review of
acrolein in support of summary information on Integrated Risk Information System (IRIS)
National Center for Environmental Assessment, Washington, DC. EPA/635/R-03/003.
Available online at: http://www.epa.gov/ncea/iris.
98 Morris JB, Symanowicz PT, Olsen JE, et al. 2003. Immediate sensory nerve-mediated
respiratory responses to irritants in healthy and allergic airway-diseased mice. J Appl Physiol
94(4):1563-1571. Docket EPA-HQ-OAR-2009-0472-0396.
99 Perera, P.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. DocketEPA-HQ-OAR-2009-0472-0372.
100 Perera, F.P.; Rauh, V.; Whyatt, R.M.; Tsai, W.Y.; Tang, D.; Diaz, D.; Hoepner, L.; Barr,
D.; Tu, Y.H.; Camann, D.; Kinney, P. (2006) Effect of prenatal exposure to airborne
polycyclic aromatic hydrocarbons on neurodevelopment in the first 3 years of life among
inner-city children. Environ Health Perspect 114: 1287-1292. Docket EPA-HQ-OAR-2009-
0472-0373.
101U. S. EPA. 2004. Toxicological Review of Naphthalene (Reassessment of the Inhalation
Cancer Risk), Environmental Protection Agency, Integrated Risk Information System,
Research and Development, National Center for Environmental Assessment, Washington,
DC. This material is available electronically at http://www.epa.gov/iris/subst/0436.htm.
Docket EPA-HQ-OAR-2009-0472-0272.
102 Oak Ridge Institute for Science and Education. (2004). External Peer Review for the IRIS
Reassessment of the Inhalation Carcinogenicity of Naphthalene. August 2004.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=84403. Docket EPA-HQ-OAR-2009-
0472-0273.
103 National Toxicology Program (NTP). (2004). llth 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.
7-159
-------
Regulatory Impact Analysis
104 International Agency for Research on Cancer (IARC). (2002). Monographs on the
Evaluation of the Carcinogenic Risk of Chemicals for Humans. Vol. 82. Lyon, France.
Docket EPA-HQ-OAR-2009-0472-0274.
105 U. S. EPA. 1998. Toxicological Review of Naphthalene, Environmental Protection
Agency, Integrated Risk Information System, Research and Development, National Center for
Environmental Assessment, Washington, DC. This material is available electronically at
http://www.epa.gov/iris/subst/0436.htm
106 Zhou, Y.; Levy, J.I. (2007) Factors influencing the spatial extent of mobile source air
pollution impacts: a meta-analysis. BMC Public Health 7: 89. doi:10.1186/1471-2458-7-89.
Docket EPA-HQ-OAR-2009-0472-11375.
107 HEI Panel on the Health Effects of Air Pollution. (2010) Traffic-related air pollution: a
critical review of the literature on emissions, exposure, and health effects. [Online at
www.healtheffects.org1. Docket EPA-HQ-OAR-2009-0472-11382.
108 Salam, M.T.; Islam, T.; Gilliland, F.D. (2008) Recent evidence for adverse effects of
residential proximity to traffic sources on asthma. Current Opin Pulm Med 14: 3-8. Docket
EPA-HQ-OAR-2009-0472-11381.
109 Holguin, F. (2008) Traffic, outdoor air pollution, and asthma. Immunol Allergy Clinics
North Am 28: 577-588.
110 Adar, S.D.; Kaufman, J.D. (2007) Cardiovascular disease and air pollutants: evaluating
and improving epidemiological data implicating traffic exposure. Inhal Toxicol 19: 135-149.
Docket EPA-HQ-OAR-2009-0472-11376.
111 Raaschou-Nielsen, O.; Reynolds, P. (2006) Air pollution and childhood cancer: a review
of the epidemiological literature. Int J Cancer 118: 2920-2929. Docket EPA-HQ-OAR-2009-
0472-11380.
112 U.S. Census Bureau (2008) American Housing Survey for the United States in 2007.
Series H-150 (National Data), Table 1A-6. [Accessed at
http://www.census.gov/hhes/www/housing/ahs/ahs07/ahs07.html on January 22, 2009]
113 Lena, T.S.; Ochieng, V.; Carter, M.; Holguin-Veras, J.; Kinney, P.L. (2002) Elemental
carbon and PM2.5 levels in an urban community heavily impacted by truck traffic. Environ
Health Perspect 110: 1009-1015. Docket EPA-HQ-OAR-2009-0472-11379.
114 Wier, M.; Sciammas, C.; Seto, E.; Bhatia, R.; Rivard, T. (2009) Health, traffic, and
environmental justice: collaborative research and community action in San Francisco,
California. Am J Public Health 99: S499-S504. Docket EPA-HQ-OAR-2009-0472-11373.
115 Forkenbrock, D.J. and L.A. Schweitzer, Environmental Justice and Transportation
Investment Policy. Iowa City: Univerity of Iowa, 1997.
116 Appatova, A.S.; Ryan, P.H.; LeMasters, O.K.; Grinshpun, S.A. (2008) Proximal exposure
of public schools and students to major roadways: a nationwide US survey. J Environ Plan
Mgmt. Docket EPA-HQ-OAR-2009-0472-11378.
7-160
-------
Environmental and Health Impacts
117 Green, R.S.; Smorodinsky, S.; Kim, J.J.; McLaughlin, R.; Ostro, B. (2004) Proximity of
California public schools to busy roads. Environ Health Perspect 112: 61-66. Docket EPA-
HQ-OAR-2009-0472-11377.
118 Houston, D.; Ong, P.; Wu, J.; Winer, A. (2006) Proximity of licensed child care facilities
to near-roadway vehicle pollution. Am J Public Health 96: 1611-1617. Docket EPA-HQ-
OAR-2009-0472-11383.
119 Wu, Y.; Batterman, S. (2006) Proximity of schools in Detroit, Michigan to automobile and
truck traffic. J Exposure Sci Environ Epidemiol 16: 457-470. Docket EPA-HQ-OAR-2009-
0472-11374.
120
U.S. EPA (2009). final PM ISA, pg 9-19 through 9-23
121 U.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. Docket EPA-HQ-
OAR-2009-0472-0343.
1 99
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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
123 Winner, W.E., and CJ. Atkinson. 1986. "Absorption of air pollution by plants, and
consequences for growth." Trends in Ecology and Evolution 7:15-18. Docket EPA-HQ-
OAR-2009-0472-0344.
124 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
125 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. Docket EPA-HQ-OAR-2009-0472-0345.
1 "Jf\
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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
127 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
128 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
1 9Q
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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
7-161
-------
Regulatory Impact Analysis
130Ollinger, 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. Docket EPA-HQ-OAR-2009-0472-0346.
131 Winner, W.E. (1994). Mechanistic analysis of plant responses to air pollution. Ecological
Applications, 4(4), 651-661. Docket EPA-HQ-OAR-2009-0472-0347.
1 ^9
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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
133 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
134 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.
135 De Steiguer, J., Pye, J., Love, C. (1990). Air Pollution Damage to U.S. Forests. Journal of
Forestry, 88(8), 17-22. Docket EPA-HQ-OAR-2009-0472-0348.
136 Pye, J.M. (1988). Impact of ozone on the growth and yield of trees: A review. Journal of
Environmental Quality, 17, 347-360. Docket EPA-HQ-OAR-2009-0472-0349.
137 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
-I Q O
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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
139 McBride, 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. Docket EPA-HQ-
OAR-2009-0472-0374.
140 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). Docket EPA-HQ-OAR-2009-
0472-0375.
141 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
142Kopp, 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. Docket EPA-HQ-OAR-2009-0472-0376.
143 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. Docket EPA-HQ-OAR-2009-0472-0397.
7-162
-------
Environmental and Health Impacts
144 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. Docket EPA-HQ-OAR-2009-0472-0398.
14 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. Docket EPA-HQ-OAR-2009-0472-0399.
146 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. Docket EPA-HQ-OAR-2009-
0472-0099, EPA-HQ-OAR-2009-0472-0100, and EPA-HQ-OAR-2009-0472-0101.
147 Grulke, N.E. (2003). The physiological basis of ozone injury assessment attributes in
Sierran conifers. In A. Bytnerowicz, MJ. 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. Docket EPA-HQ-OAR-2009-0472-0400.
148 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. Docket EPA-HQ-OAR-2009-0472-1200.
149 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. Docket EPA-HQ-OAR-
2009-0472-1201.
150White, 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. Docket EPA-HQ-OAR-2009-0472-1200.
151 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. Docket EPA-HQ-OAR-
2009-0472-1201.
152 Coulston, J.W., Riitters, K.H., Smith, G.C. (2004). A preliminary assessment of the
Montreal process indicators of air pollution for the United States. Environmental Monitoring
and Assessment, 95, 57-74. Docket EPA-HQ-OAR-2009-0472-1202.
1 3 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants.
EPA/600/R-05/004aF-cF. Docket EPA-HQ-OAR-2009-0472-0099, EPA-HQ-OAR-2009-
0472-0100, and EPA-HQ-OAR-2009-0472-0101.
154 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. Docket EPA-HQ-OAR-
2009-0472-1201.
1 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants.
EPA/600/R-05/004aF-cF. Docket EPA-HQ-OAR-2009-0472-0099, EPA-HQ-OAR-2009-
0472-0100, and EPA-HQ-OAR-2009-0472-0101.
7-163
-------
Regulatory Impact Analysis
156 US 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. Docket EPA-HQ-OAR-2009-0472-0106.
157 Chappelka, A.H., Samuelson, LJ. (1998). Ambient ozone effects on forest trees of the
eastern United States: a review. New Phytologist, 139,91-108. Docket EPA-HQ-OAR-
2009-0472-1203
158 U.S. EPA (2009) Final PM ISA
159 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. Docket EPA-HQ-OAR-2009-0472-0098.
160 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. Docket EPA-
HQ-OAR-2009-0472-0014.
161 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. Docket EPA-HQ-OAR-2009-0472-0095.
162 Fenn, M.E. and Blubaugh, TJ. (2005) Winter Deposition of Nitrogen and Sulfur in the
Eastern Columbia River Gorge National Scenic Area, USDA Forest Service. Docket EPA-
HQ-OAR-2009-0472-0112.
163 Galloway, J. N.; Cowling, E. B. (2002). Reactive nitrogen and the world: 200 years of
change. Ambio 31: 64-71. Docket EPA-HQ-OAR-2009-0472-0377.
164Bricker, 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. Docket EPA-HQ-OAR-2009-0472-0378 and
EPA-HQ-OAR-2009-0472-0378.1.
165 Smith, W.H. 1991. "Air pollution and Forest Damage." Chemical Engineering News,
69(45): 30-43. Docket EPA-HQ-OAR-2009-0472-0275.
166Gawel, I.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.
Docket EPA-HQ-OAR-2009-0472-0276.
167Cotrufo, 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. Docket EPA-HQ-OAR-2009-0472-1205.
168 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. Docket EPA-HQ-OAR-2009-0472-1206.
169 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section
9.4.5.2. Docket EPA-HQ-OAR-2009-0472-11295
7-164
-------
Environmental and Health Impacts
170 Mason, R.P. and Sullivan, K.A. 1997. "Mercury in Lake Michigan." Environmental
Science & Technology, 31: 942-947. (from Delta Report "Atmospheric deposition of toxics
to the Great Lakes"). Docket EPA-HQ-OAR-2009-0472-1207.
171 Landis, M.S. and Keeler, GJ. 2002. "Atmospheric mercury deposition to Lake Michigan
during the Lake Michigan Mass Balance Study." Environmental Science & Technology, 21:
4518-24. Docket EPA-HQ-OAR-2009-0472-1208.
172 U.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. Docket EPA-HQ-OAR-2009-0472-0091.
173 National Science and Technology Council (NSTC) 1999. "The Role of Monitoring
Networks in the Management of the Nation's Air Quality." Docket EPA-HQ-OAR-2009-
0472-1209.
174 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. Docket EPA-HQ-OAR-2009-0472-
1895.
175 Rice, K.C. 1999. "Trace Element Concentrations in Streambed Sediment Across the
Conterminous United States." Environmental Science & Technology, 33: 2499-2504. Docket
EPA-HQ-OAR-2009-0472-1896.
176 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. Docket EPA-HQ-OAR-2009-0472-1897.
177 U.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. Docket EPA-HQ-OAR-2009-0472-1908.
178 U.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. Docket EPA-HQ-OAR-2009-0472-1908.
179 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. Docket EPA-HQ-OAR-2009-0472-
1909.
180 Simcik, M.F.; Eisenreich, S.J.; and Lioy, PJ. 1999. "Source apportionment and source/sink
relationship of PAHs in the coastal atmosphere of Chicago and Lake Michigan." Atmospheric
Environment, 33: 5071-5079. Docket EPA-HQ-OAR-2009-0472-1910.
181 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. Docket EPA-HQ-OAR-2009-0472-0256.
182Park, 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. Docket EPA-HQ-OAR-2009-0472-0352.
7-165
-------
Regulatory Impact Analysis
183 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. Docket EPA-HQ-OAR-2009-0472-0354.
184Arzayus, 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. Docket EPA-HQ-OAR-2009-0472-0256.
185 U.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. Docket EPA-HQ-OAR-2009-0472-0091.
186 Van Metre, P.C.; Mahler, B.J.; and Furlong, E.T. 2000. "Urban Sprawl Leaves its PAH
Signature." Environmental Science & Technology, 34: 4064-4070. Docket EPA-HQ-OAR-
2009-0472-0355.
187 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. Docket EPA-HQ-OAR-2009-0472-0359.
188 Tuhackova, J. et al. (2001) Hydrocarbon deposition and soil microflora as affected by
highway traffic. Environmental Pollution, 113: 255-262. Docket EPA-HQ-OAR-2009-0472-
0356.
189 U.S. EPA. 1991. Effects of organic chemicals in the atmosphere on terrestrial plants.
EPA/600/3-91/001. Docket EPA-HQ-OAR-2009-0472-0401.
190 Cape JN, ID Leith, J Binnie, J Content, M Donkin, M Skewes, DN Price AR Brown, AD
Sharpe. 2003. Effects of VOCs on herbaceous plants in an open-top chamber experiment.
Environ. Pollut. 124:341-343. Docket EPA-HQ-OAR-2009-0472-0357.
191 Cape JN, ID Leith, J Binnie, J Content, M Donkin, M Skewes, DN Price AR Brown, AD
Sharpe. 2003. Effects of VOCs on herbaceous plants in an open-top chamber experiment.
Environ. Pollut. 124:341-343. Docket EPA-HQ-OAR-2009-0472-0357.
192 Viskari E-L. 2000. Epicuticular wax of Norway spruce needles as indicator of traffic
pollutant deposition. Water, Air, and Soil Pollut. 121:327-337. Docket EPA-HQ-OAR-2009-
0472-1128.
193 Ugrekhelidze D, F Korte, G Kvesitadze. 1997. Uptake and transformation of benzene and
toluene by plant leaves. Ecotox. Environ. Safety 37:24-29. Docket EPA-HQ-OAR-2009-
0472-1142.
194 Kammerbauer H, H Selinger, R Rommelt, A Ziegler-Jons, D Knoppik, B Hock. 1987.
Toxic components of motor vehicle emissions for the spruce Pciea abies. Environ. Pollut.
48:235-243. Docket EPA-HQ-OAR-2009-0472-0358.
195 Byun, D.W., Ching, J. K.S. (1999). Science algorithms of the EPA models-3 community
multiscale air quality (CMAQ) modeling system. Washington, DC: U.S. Environmental
Protection Agency, Office of Research and Development. Docket EPA-HQ-OAR-2009-0472-
1915
7-166
-------
Environmental and Health Impacts
196 Byun, D.W., 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. Journal of Applied Mechanics Reviews, 59(2), 51-77. Docket
EPA-HQ-O AR-2009 -0472-1916
197 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. Docket EPA-HQ-OAR-2009-0472-1917
198 Hogrefe, C., Biswas, J., Lynn, B., Civerolo, K., Ku, J.Y., Rosenthal, J., et al. (2004).
Simulating regional-scale ozone climatology over the eastern United States: model evaluation
results. Atmospheric Environment, 38(17), 2627-2638. Docket EPA-HQ-OAR-2009-0472-
11350
199 Lin, M., Oki, T., Holloway, T., Streets, D.G., Bengtsson, M., Kanae, S. (2008). Long-
range transport of acidifying substances in East Asia-Part I: Model evaluation and sensitivity
studies. Atmospheric Environment, 42(24), 5939-5955. Docket EPA-HQ-OAR-2009-0472-
11341
200 U.S. Environmental Protection Agency. (2008). Technical support document for the final
locomotive/marine rule: Air quality modeling analyses. Research Triangle Park, N.C.: U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, Air Quality
Assessment Division. Docket EPA-HQ-OAR-2009-0472-11329
201 Grell, G., Dudhia, J., Stauffer, D. (1994). A Description of the Fifth-Generation Penn
State/NCAR Mesoscale Model (MM5), NCAR/TN-398+STR., 138 pp, National Center for
Atmospheric Research, Boulder CO. Docket EPA-HQ-OAR-2009-0472-2104
202 Grell, G., Dudhia, J., Stauffer, D. (1994). A Description of the Fifth-Generation Penn
State/NCAR Mesoscale Model (MM5), NCAR/TN-398+STR., 138 pp, National Center for
Atmospheric Research, Boulder CO. Docket EPA-HQ-O AR-2009-0472-2104
203 Byun, D.W., Ching, J. K.S. (1999). Science algorithms of EPA Models-3 Community
Multiscale Air Quality (CMAQ) modeling system, EPA/600/R-99/030, Office of Research and
Development). Please also see: http://www.cmascenter.org/. Docket EPA-HQ-OAR-2009-
0472-1915
204 Le Sager, P. Yantosca, B., Carouge, C. (2008). GEOS-CHEMv8-01-02 User's Guide,
Atmospheric Chemistry Modeling Group, Harvard University, Cambridge, MA, December
18, 2008. Docket EPA-HQ-OAR-2009-0472-11344
205 U.S. EPA, (2004), Procedures for Estimating Future PM2.5 Values for the CAIR Final
Rule by Application of the (Revised) Speciated Modeled Attainment Test (SMAT)- Updated
11/8/04. Docket EPA-HQ-OAR-2009-0472-11327
or\/C
U.S. EPA, (2008), Control of Emissions from Nonroad Spark-Ignition Engines and
Equipment, Technical Support Document. EPA 454/R-08-005. Docket EPA-HQ-OAR-2009-
0472-11330
207 U.S. EPA (2007) Guidance on the Use of Models and Other Analyses For Demonstrating
Attainment of Air Quality Goals for Ozone, PM2.5, and Regional Haze; EPA-454/B-07-002;
Research Triangle Park, NC; April 2007. Docket EPA-HQ-OAR-2009-0472-11328
7-167
-------
Regulatory Impact Analysis
208 Yarwood G, Rao S, Yocke M, Whitten GZ (2005) Updates to the Carbon Bond Chemical
Mechanism: CB05. Final Report to the US EPA, RT-0400675, December 8, 2005.
http://www.camx.com/publ/pdfs/CB05_Final_Report_120805.pdf. Docket EPA-HQ-OAR-
2009-0472-0111
209 Dodge, M.C., 2000. Chemical oxidant mechanisms for air quality modeling: critical
review. Atmospheric Environment 34, 2103-2130. Docket EPA-HQ-OAR-2009-0472-11338
210 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
211 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
212 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
213 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
214 Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat,
O.K., Ravishankara, A.R., Kolb, C.E., Molina, M.J., Finlayson-Pitts, B.J. (2003) Chemical
Kinetics and Photochemical Data for use in Atmospheric Studies, Evaluation Number 14.
NASA Jet Propulsion Laboratory http://jpldataeval.jpl.nasa.gov/index.html. Docket EPA-
HQ-OAR-2009-0472-11320
215 Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat,
O.K., Ravishankara, A.R., Kolb, C.E., Molina, M.J., Finlayson-Pitts, B.J. (2003) Chemical
Kinetics and Photochemical Data for use in Atmospheric Studies, Evaluation Number 14.
NASA Jet Propulsion Laboratory http://jpldataeval.jpl.nasa.gov/index.html. Docket EPA-HQ-
OAR-2009-0472-11320
216 Yarwood, G., Rao, S., Yocke, M., Whitten, G.Z., 2005. Updates to the Carbon Bond
Mechanism: CB05. Final Report to the US EPA, RT-0400675. Yocke and Company, Novato,
CA. Docket EPA-HQ-OAR-2009-0472-0111
7-168
-------
Environmental and Health Impacts
217 Luecken, D.J., Phillips, S., Sarwar, G., Jang, C., 2008b. Effects of using the CB05 vs.
SAPRC99 vs. CB4 chemical mechanism on model predictions: Ozone and gas-phase
photochemical precursor concentrations. Atmospheric Environment 42, 5805-5820. Docket
EPA-HQ-OAR-2009-0472-11343
218 Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat,
O.K., Ravishankara, A.R., Kolb, C.E., Molina, M.J., Finlayson-Pitts, B.J., 2003. Chemical
Kinetics and Photochemical Data for use in Atmospheric Studies, Evaluation Number 14.
NASA Jet Propulsion Laboratory. Docket EPA-HQ-OAR-2009-0472-11320
219 Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat,
O.K., Ravishankara, A.R., Kolb, C.E., Molina, M.J., Finlayson-Pitts, B.J., 2003. Chemical
Kinetics and Photochemical Data for use in Atmospheric Studies, Evaluation Number 14.
NASA Jet Propulsion Laboratory. Docket EPA-HQ-OAR-2009-0472-11320
220 Atkinson R, Arey J (2003) Atmospheric Degradation of Volatile Organic Compounds.
Chem Rev 103: 4605-4638. Docket EPA-HQ-OAR-2009-0472-11334
221 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
222 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
223 Edney, E. O., T. E. Kleindienst, M. Lewandowski, and J. H. Offenberg, 2007. Updated
SOA chemical mechanism for the Community Multi-Scale Air Quality model, EPA 600/X-
07/025, U.S. EPA, Research Triangle Park, NC. Docket EPA-HQ-OAR-2009-0472-11345
224 Carlton, A.G., B. J. Turpin, K. Altieri, S. Seitzinger, R. Mathur, S. Roselle, R. J. Weber,
(2008), CMAQ model performance enhanced when in-cloud SOA is included: comparisons
of OC predictions with measurements, Environ. Sci. Technol. 42 (23), 8798-8802. Docket
EPA-HQ-OAR-2009-0472-11336
225 Lewandowski M, M Jaoui, JH Offenberg , TE Kleindienst, EO Edney, RJ Sheesley, JJ
Schauer (2008) Primary and secondary contributions to ambient PM in the midwestern
United States, Environ Sci Technol 42(9):3303-3309. http://pubs.acs.org/cgi-
bin/article.cgi/esthag/2008/42/i09/html/es0720412.html. Docket EPA-HQ-OAR-2009-0472-
11340
226 Kleindienst TE, M Jaoui, M Lewandowski, JH Offenberg, EO Edney (2007) Estimates of
the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at
7-169
-------
Regulatory Impact Analysis
a southeastern U.S. location, Atmos Environ 41(37):8288-8300. Docket EPA-HQ-OAR-
2009-0472-11356
227 Offenberg JH, CW Lewis, M Lewandowski, M Jaoui, TE Kleindienst, EO Edney (2007)
Contributions of Toluene and D-pinene to SOA Formed in an Irradiated Toluene/D-
pinene,NOx/Air Mixture: Comparison of Results Using 14C Content and SOA Organic Tracer
Methods, Environ Sci Technol 41: 3972-3976. Docket EPA-HQ-OAR-2009-0472-11318
228 Pandis, S.N., Harley, R.A., Cass, G.R., Seinfeld, J.H. (1992) Secondary organic aerosol
formation and transport. Atmos Environ 26, 2269-2282. Docket EPA-HQ-OAR-2009-0472-
11319
229 Takekawa, H. Minoura, H. Yamazaki, S. (2003) Temperature dependence of secondary
organic aerosol formation by photo-oxidation of hydrocarbons. Atmos Environ 37: 3413-
3424. Docket EPA-HQ-OAR-2009-0472-11358
230 Kleeman, M.J., Ying, Q., Lu, J., Mysliwiec, M.J., Griffin, R.J., Chen, J., Clegg, S. (2007)
Source apportionment of secondary organic aerosol during a severe photochemical smog
episode. Atmos Environ 41: 576-591. Docket EPA-HQ-OAR-2009-0472-11313
231 Robinson, A. L.; Donahue, N. M.; Shrivastava, M.; Weitkamp, E. A.; Sage, A. M.;
Grieshop, A. P.; Lane, T. E.; Pierce, J. R.; Pandis, S. N. (2007) Rethinking organic aerosol:
Semivolatile emissions and photochemical aging. Science 315: 1259-1262. Docket EPA-HQ-
OAR-2009-0472-11357
232 Byun DW, Schere, KL (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: 51-76. Docket EPA-HQ-
OAR-2009-0472-1916
233 U. S. EPA (2002) National Air Quality and Emissions Trends Report, 2001. EPA 454/K-
02-001, September 2002. http://www.epa.gov/air/airtrends/aqtrnd01/summary.pdf. Docket
EPA-HQ-OAR-2009-0472-0092
234 U. S. EPA. (2009) 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata2002/risksum.html. Docket EPA-HQ-OAR-2009-0472-
11321
235 U. S. EPA (2007) Regulatory Impact Analysis for the Control of Hazardous Air Pollutants
from Mobile Sources Rule, Chapter 3, Air Quality and Resulting Health and Welfare Effects
of Air Pollution from Mobile Sources. 72 FR 8428, February 26, 2007.
http://www.epa.gov/otaq/regs/toxics/420r07002.pdf. Docket EPA-HQ-OAR-2009-0472-0107
236 Kleindienst, T.E. (2008) Hypothetical SOA Production from Ethanol Photooxidation.
Memo to the Docket EPA-HQ-OAR-2005-0161. Docket EPA-HQ-OAR-2009-0472-11314
7-170
-------
Environmental and Health Impacts
237 Turpin, B.J., Huntzicker, J.J., Larson, S.M., Cass, G.R. (1991) Los Angeles Summer
Midday Participate Carbon: Primary and Secondary Aerosol. Environ Sci Technol 25: 1788-
1793. Docket EPA-HQ-OAR-2009-0472-11359
238 Turpin, B.J., Huntzicker, JJ. (1995) Identification of Secondary Organic Aerosol Episodes
and Quantitation of Primary and Secondary Organic Aerosol Concentrations During SCAQS.
Atmos Environ 29(23): 3527-3544. Docket EPA-HQ-OAR-2009-0472-11360
239 Bae M-S, Schauer JJ, Turner JR (2006) Estimation of the Monthly Average Ratios of
Organic Mass to Organic Carbon for Fine Particulate Matter at an Urban Site, Aerosol Sci
Technol 40(12): 1123-1139. http://dx.doi.org/10.1080/02786820601004085. Docket EPA-
HQ-OAR-2009-0472-11335
240 Kleindienst TE, M Jaoui, M Lewandowski, JH Offenberg, EO Edney (2007) Estimates of
the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at
a southeastern U.S. location. Atmos Environ 41(37):8288-8300. Docket EPA-HQ-OAR-
2009-0472-11356
241 Offenberg JH, CW Lewis, M Lewandowski, M Jaoui, TE Kleindienst, EO Edney (2007)
Contributions of Toluene and D-pinene to SOA Formed in an Irradiated Toluene/D-
pinene,NOx/Air Mixture: Comparison of Results Using 14C Content and SOA Organic Tracer
Methods, Environ Sci Technol 41: 3972-3976. Docket EPA-HQ-OAR-2009-0472-11318
242 Claeys M, R Szmigielski, I Kourtchev, P Van der Veken, R Vermeylen, W Maenhaut, M
Jaoui, TE Kleindienst, M Lewandowski, JH Offenberg, EO Edney (2007)
Hydroxydicarboxylic acids: Markers for secondary organic aerosol from the photooxidation
of D-pinene. Environ Sci Technol 41(5): 1628-1634. Docket EPA-HQ-OAR-2009-0472-
11337
243 Edney EO, TE Kleindienst, M Jaoui, M Lewandowski, JH Offenberg, W Wang, M Claeys
(2005) Formation of 2-methyl tetrols and 2-methylglyceric acid in secondary organic aerosol
from laboratory irradiated isoprene/NOx/SOi/air mixtures and their detection in ambient
PM2.5 samples collected in the Eastern United States. Atmos Environ 39: 5281-5289. Docket
EPA-HQ-OAR-2009-0472-11339
244 Jaoui M, TE Kleindienst, M Lewandowski, JH Offenberg, EO Edney (2005) Identification
and quantification of aerosol polar oxygenated compounds bearing carboxylic or hydroxyl
groups. 2. Organic tracer compounds from monoterpenes. Environ Sci Technol 39: 5661-
5673. Docket EPA-HQ-OAR-2009-0472-11352
245 Kleindienst TE, TS Conver, CD Mclver, EO Edney (2004) Determination of secondary
organic aerosol products from the photooxidation of toluene and their implications in ambient
PM2.5. J Atmos Chem 47: 70-100. Docket EPA-HQ-OAR-2009-0472-11315
246 Kleindienst TE, TS Conver, CD Mclver, EO Edney (2004) Determination of secondary
organic aerosol products from the photooxidation of toluene and their implication in ambient
PM2.5, J Atmos Chem 47: 70-100. Docket EPA-HQ-OAR-2009-0472-11315
7-171
-------
Regulatory Impact Analysis
247 Izumi L, T Fukuyama (1990) Photochemical aerosol formation from aromatic
hydrocarbons in the presence of NOX, Atmos Environ 24A: 1433. Docket EPA-HQ-OAR-
2009-0472-11351
248 Martin-Reviego M, K Wirtz (2005) Is benzene a precursor for secondary organic aerosol?
Environ Sci Technol 39: 1045-1054. Docket EPA-HQ-OAR-2009-0472-11343
249 Ng, N. L., Kroll, J. H., Chan, A. W. H., Chabra, P. S., Flagan, R. C., Seinfield, J. H.,
Secondary organic aerosol formation from m-xylene, toluene, and benzene, Atmospheric
Chemistry and Physics Discussion, 7, 3909-3922, 2007. Docket EPA-HQ-OAR-2009-0472-
11316
250 Lewandowski M, M Jaoui, JH Offenberg , TE Kleindienst, EO Edney, RJ Sheesley, JJ
Schauer (2008) Primary and secondary contributions to ambient PM in the midwestern
United States, Environ Sci Technol 42(9):3303-3309. http://pubs.acs.org/cgi-
bin/article.cgi/esthag/2008/42/i09/html/es0720412.html. Docket EPA-HQ-OAR-2009-0472-
11340
251 Kleindienst TE, M Jaoui, M Lewandowski, JH Offenberg, EO Edney (2007) Estimates of
the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at
a southeastern U.S. location. Atmos Environ 41(37):8288-8300. Docket EPA-HQ-OAR-
2009-0472-11356
252 Henze DK, JH Seinfeld (2006) Global secondary organic aerosol from isoprene oxidation.
Geophys Res Lett 33: L09812. doi:10.1029/2006GL025976. Docket EPA-HQ-OAR-2009-
0472-11347
253 Henze DK.;J Seinfeld (2006) Global secondary organic aerosol from isoprene oxidation.
Geophys. Res. Lett 33 L09812. doi:10.1029/2006GL025976. Docket EPA-HQ-OAR-2009-
0472-11347
254 Jaoui M, M Lewandowski, TE Kleindienst, JH Offenberg, EO Edney (2007) p -
Caryophyllinic acid: An atmospheric tracer for p-caryophyllene secondary organic aerosol.
Geophys Res Lett 34: L05816. doi:10.1029/2006GL028827. Docket EPA-HQ-OAR-2009-
0472-1917
255 Griffin RJ, DR Cocker III, RC Flagan, JH Seinfeld (1999) Organic aerosol formation from
oxidation of biogenic hydrocarbons, J Geophys Res 104: 3555-3567. Docket EPA-HQ-OAR-
2009-0472-11346
256 Ng NL, JH Kroll, AWH Chan, PS Chabra, RC Flagan, JH Seinfeld (2007) Secondary
organic aerosol formation from m-xylene, toluene, and benzene. Atmos Chem Phys 7: 3909-
3922. Docket EPA-HQ-OAR-2009-0472-11316
257 Hildebrandtl, L., Donahuel, N. M, Pandisl, S. N. (2009) High formation of secondary
organic aerosol from the photo-oxidation of toluene. Atmos Chem Phys 9: 2973-2986.
Docket EPA-HQ-OAR-2009-0472-11349
7-172
-------
Environmental and Health Impacts
258 Atkinson, R., Baulch, D.L., Cox, R.A., Crowley, J.N., Hampson, R.F. Jr., Hynes, R.G.,
Jenkin, M.E., Kerr, J.A., Rossi, M.J., Troe, J. (2005) Evaluated Kinetic and Photochemical
Data for Atmospheric Chemistry - IUPAC Subcommittee on Gas Kinetic Data Evaluation for
Atmospheric Chemistry. July 2005 web version, http://www.mpac-
kinetic.ch.cam.ac.uk/index.html. Docket EPA-HQ-OAR-2009-0472-11333
259 Sander, S.P., Friedl, R.R., Golden, D.M., Kurylo, M.J., Huie, R.E., Orkin, V.L., Moortgat,
O.K., Ravishankara, A.R., Kolb, C.E., Molina, M.J., Finlayson-Pitts, B.J. (2003) Chemical
Kinetics and Photochemical Data for use in Atmospheric Studies, Evaluation Number 14.
NASA Jet Propulsion Laboratory, http://jpldataeval.jpl.nasa.gov/index.html. Docket EPA-
HQ-OAR-2009-0472-11320
260 Finlayson-Pitts BJ, Pitts JN Jr. (1986) Atmospheric Chemistry: Fundamentals and
Experimental Techniques, Wiley, New York.
261 Yarwood G, Rao S, Yocke M, Whitten GZ (2005) Updates to the Carbon Bond Chemical
Mechanism: CB05. Final Report to the US EPA, RT-0400675, December 8, 2005.
http://www.camx.com/publ/pdfs/CB05_Final_Report_120805.pdf. Docket EPA-HQ-OAR-
2009-0472-0111
262http://www.cmascenter.org/help/model_docs/cmaq/4.7/RELEASE_NOTES.txt
263 U.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.
264 PM Standards Revision - 2006: Timeline. Retrieved on March 19, 2009 from
http://www.epa.gov/oar/particlepollution/naaqsrev2006.htmltttimeline
265 EPA 2010, Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis. EPA-
420-R-10-006. February 2010. Docket EPA-HQ-OAR-2009-0472-11332. see also 75 FR
14670, March 26, 2010.
266 EPA 2010, Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis. EPA-
420-R-10-006. February 2010. Sections 3.4.2.1.2 and 3.4.3.3. Docket EPA-HQ-OAR-2009-
0472-11332
267 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.
268 U. S. Environmental Protection Agency (2007). Control of Hazardous Air Pollutants from
Mobile Sources; Final Rule. 72 FR 8434, February 26, 2007. Docket EPA-HQ-OAR-2009-
0472-0271, 0271.1 and 0271.2
7-173
-------
Regulatory Impact Analysis
269 U. S. Environmental Protection Agency (2007). Control of Hazardous Air Pollutants from
Mobile Sources; Final Rule. 72 FR 8434, February 26, 2007. Docket EPA-HQ-OAR-2009-
0472-0271,0271.1 and 0271.2
270 U. S. EPA. (2009) 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata2002/. Docket EPA-HQ-OAR-2009-0472-11321
271 U. S. EPA. (2009) 2002 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/nata2002/risksum.html. Docket EPA-HQ-OAR-2009-0472-
11322
272 U.S. EPA. U.S. EPA's 2008 Report on the Environment (Final Report). U.S.
Environmental Protection Agency, Washington, D.C., EPA/600/R-07/045F (NTIS PB2008-
112484). Docket EPA-HQ-OAR-2009-0472-11298. Updated data available online at:
http://cfpub.epa.gov/eroe/index.cfm?fuseaction=detail.viewlnd&ch=46&subtop=341&lv=list.
listByChapter&r=201744
273 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.
Docket EPA-HQ-OAR-2009-0472-0096
274 U.S. Environmental Protection Agency. (2006). Final Regulatory Impact Analysis (RIA)
for the Proposed National Ambient Air Quality Standards for Particulate Matter. Prepared
by: Office of Air and Radiation. Retrieved March, 26, 2009 at
http://www.epa.gov/ttn/ecas/ria.html. EPA-HQ-OAR-2009-0472-0240
275 U.S. Environmental Protection Agency. (2008). Final Ozone NAAQS Regulatory Impact
Analysis. Prepared by: Office of Air and Radiation, Office of Air Quality Planning and
Standards. Retrieved March, 26, 2009 at http://www.epa.gov/ttn/ecas/ria.html. EPA-HQ-
OAR-2009-0472-0238
276 U.S. Environmental Protection Agency (U.S. EPA). 2009. Regulatory Impact Analysis:
National Emission Standards for Hazardous Air Pollutants from the Portland Cement
Manufacturing Industry. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. April. Available on the Internet at
. EPA-HQ-OAR-
2009-0472-0241
277 U.S. Environmental Protection Agency (U.S. EPA). 2010. Final NO2 NAAQS Regulatory
Impact Analysis (RIA). Office of Air Quality Planning and Standards, Research Triangle
Park, NC. April. Available on the Internet at
http://www.epa.gov/ttn/ecas/regdata/RIAs/FinalNO2RIAfulldocument.pdf. Accessed March
15. EPA-HQ-OAR-2009-0472-0237
278 U.S. Environmental Protection Agency. 2009. Regulatory Impact Analysis: Control of
Emissions of Air Pollution from Category 3 Marine Diesel Engines. EPA-420-R-09-019,
December 2009. Prepared by Office of Air and Radiation.
7-174
-------
Environmental and Health Impacts
http://www.epa.gov/otaq/regs/nonroad/marine/ci/420r09019.pdf. Accessed February 9, 2010.
EPA-HQ-OAR-2009-0472-0283
279 Science Advisory Board. 2001. NATA - Evaluating the National-Scale Air Toxics
Assessment for 1996 - an SAB Advisory, http://www.epa.gov/ttn/atw/sab/sabrev.html. EPA-
HQ-OAR-2009-0472-0244
280 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. EPA-HQ-OAR-2009-0472-
0382
281 GeoLytics Inc. (2002). Geolytics CensusCD® 2000 Short Form Blocks. CD-ROM
Release 1.0. GeoLytics, Inc. East Brunswick, NJ. Available: http://www.geolytics.com/
[accessed 29 September 2004]
282 Woods & Poole Economics Inc. 2008. Population by Single Year of Age CD. CD-ROM.
Woods & Poole Economics, Inc. Washington, D.C. EPA-HQ-OAR-2009-0472-0011
283 U.S. Environmental Protection Agency. (2006). Air quality criteria for ozone and related
photochemical oxidants (second external review draft). Research Triangle Park, NC:
National Center for Environmental Assessment; report no. EPA/600R-05/004aB-cB, 3v.
Available: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=137307[March 2006] EPA-
HQ-OAR-2009-0472-0099,EPA-HQ-OAR-2009-0472-0100, EPA-HQ-OAR-2009-0472-
0101
284 U.S. Environmental Protection Agency, 2004. Air Quality Criteria for Particulate Matter
Volume II of II. National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
EPA/600/P-99/002bF. EPA-HQ-OAR-2009-0472-0097
285 World Health Organization (WHO). (2003). Health Aspects of Air Pollution with
Particulate Matter, Ozone and Nitrogen Dioxide: Report on a WHO Working Group. World
Health Organization. Bonn, Germany. EUR/03/5042688.
286 Anderson HR, Atkinson RW, Peacock JL, Marston L, Konstantinou K. (2004). Meta-
analysis of time-series studies and panel studies of Particulate Matter (PM) and Ozone (O3):
Report of a WHO task group. Copenhagen, Denmark: World Health Organization.
287 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. EPA-HQ-OAR-2009-0472-1662
288 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.
EPA-HQ-OAR-2009-0472-0233
289 Schwartz, J. (2005) How sensitive is the association between ozone and daily deaths to
control for temperature? Am. J. Respir. Crit. Care Med. Ill: 627-631. EPA-HQ-OAR-2009-
0472-1678
7-175
-------
Regulatory Impact Analysis
290 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. EPA-HQ-OAR-2009-0472-0222
291 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'. EPA-HQ-OAR-2009-
0472-0231
292 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. EPA-HQ-OAR-2009-
0472-0236
293 Pope, C.A., III, R.T. Burnett, MJ. Thun, E.E. Calle, D. Krewski, K. Ito, and G.D.
Thurston. (2002). "Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to
Fine Particulate Air Pollution." Journal of the American Medical Association 287:1132-1141.
EPA-HQ-OAR-2009-0472-0263
294 Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery. (2006). Reduction in Fine
Particulate Air Pollution and Mortality. American Journal of Respiratory and Critical Care
Medicine. 173: 667-672. EPA-HQ-OAR-2009-0472-1661
295 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. EPA-HQ-OAR-
2009-0472-0242
296 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. EPA-HQ-OAR-2009-0472-0382
297 Abbey, D.E., B.L. Hwang, RJ. 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.
EPA-HQ-OAR-2009-0472-0432
298 Peters, A., D.W. Dockery, I.E. Muller, and M.A. Mittleman. (2001). Increased Particulate
Air Pollution and the Triggering of Myocardial Infarction. Circulation. 103:2810-2815. EPA-
HQ-OAR-2009-0472-0239
299 Schwartz J. (1995). Short term fluctuations in air pollution and hospital admissions of the
elderly for respiratory disease. Thorax. 50(5):531-538.
300 Schwartz J. (1994a). PM(10) Ozone, and Hospital Admissions For the Elderly in
Minneapolis St Paul, Minnesota. Arch Environ Health. 49(5):366-374. EPA-HQ-OAR-2009-
0472-1673
301 Schwartz J. (1994b). Air Pollution and Hospital Admissions For the Elderly in Detroit,
Michigan. AmJRespir Crit Care Med. 150(3):648-655. EPA-HQ-OAR-2009-0472-1674
302 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. EPA-HQ-OAR-2009-0472-1673
7-176
-------
Environmental and Health Impacts
303 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. EPA-HQ-OAR-2009-0472-0223
304 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 and Health. Special Report. Boston, MA: Health Effects Institute.
305 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
and Health. Special Report. Health Effects Institute, Boston, MA. EPA-HQ-OAR-2009-0472-
1674
306 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. EPA-HQ-OAR-2009-0472-1664
307 Sheppard, L. (2003). Ambient Air Pollution and Nonelderly Asthma Hospital Admissions
in Seattle, Washington, 1987-1994. In Revised Analyses of Time-Series Studies of Air
Pollution and Health. Special Report. Boston, MA: Health Effects Institute. EPA-HQ-
OAR-2009-0472-0318
308 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. EPA-HQ-
OAR-2009-0472-0234
309 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. EPA-HQ-OAR-2009-0472-
1663
310 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. EPA-HQ-OAR-2009-0472-0246
311 Norris, 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. EPA-HQ-OAR-
2009-0472-0318
312 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. EPA-HQ-OAR-2009-0472-0225
313 Pope, C.A., III, D.W. Dockery, J.D. Spengler, and M.E. Raizenne. (1991). Respiratory
Health and PMio Pollution: A Daily Time Series Analysis. American Review of Respiratory
Diseases 144:668-674. EPA-HQ-OAR-2009-0472-1672
314 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.
7-177
-------
Regulatory Impact Analysis
315 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.
316 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. EPA-HQ-OAR-2009-0472-1671
317 Ostro, B.D. (1987). Air Pollution and Morbidity Revisited: A Specification Test. Journal
of Environmental Economics Management 14:87-98. EPA-HQ-OAR-2009-0472-1670
318 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. EPA-HQ-OAR-2009-0472-1675
319 Chen L, Jennison BL, Yang W, Omaye ST. (2000). Elementary school absenteeism and
air pollution. Inhal Toxicol 12(11):997-1016. EPA-HQ-OAR-2009-0472-0224
320 Ostro, B.D. and S. Rothschild. (1989). Air Pollution and Acute Respiratory Morbidity:
An Observational Study of Multiple Pollutants. Environmental Research 50:238-247. EPA-
HQ-OAR-2009-0472-0364
321 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. EPA-HQ-OAR-2009-0472-4664
322 National Research Council (NRC). (2002). Estimating the Public Health Benefits of
Proposed Air Pollution Regulations. Washington, DC: The National Academies Press.
323 Abt Associates, Inc. October 2005. Methodology for County-level Mortality Rate
Projections. Memorandum to Bryan Hubbell and Zachary Pekar, U.S. EPA.
324 U.S. Environmental Protection Agency. (2009). Regulatory Impact Analysis: National
Emission Standards for Hazardous Air Pollutants from the Portland Cement Manufacturing
Industry. Office of Air and Radiation. Retrieved on May 4, 2009, from
http://www.epa.gov/ttn/ecas/regdata/RIAs/portlandcementria 4-20-09.pdf EPA-HQ-OAR-
2009-0472-0241
325 Viscusi, W.K., W.A. Magat, and J. Huber. (1991). Pricing Environmental Health Risks:
Survey Assessments of Risk-Risk and Risk-Dollar Trade-Offs for Chronic Bronchitis.
Journal of Environmental Economics and Management 21:32-51.
326 Cropper, M.L., and AJ. Krupnick. (1990). The Social Costs of Chronic Heart and Lung
Disease. Resources for the Future. Washington, DC. Discussion Paper QE 89-16-REV.
327 Russell, M.W., D.M. Huse, S. Drowns, E.G. 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. EPA-HQ-OAR-2009-0472-1666
328 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. EPA-HQ-
OAR-2009-0472-1669
7-178
-------
Environmental and Health Impacts
329 Agency for Healthcare Research and Quality (AHRQ). (2000). HCUPnet, Healthcare
Cost and Utilization Project. Rockville, MD. Accessed April 10, 2009, from
http ://hcupnet. ahrq. gov/
330 Smith, D.H., D.C. Malone, K.A. Lawson, LJ. 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 Pt l):787-793. EPA-HQ-OAR-2009-0472-
1667
331 Stanford, R., T. McLaughlin, and LJ. Okamoto. (1999). The Cost of Asthma in the
Emergency Department and Hospital. American Journal of Respiratory and Critical Care
Medicine 160(1):211-215. EPA-HQ-OAR-2009-0472-1668
332 Industrial Economics, Incorporated (lEc). (1994). Memorandum to Jim DeMocker,
Office of Air and Radiation, Office of Policy Analysis and Review, U.S. Environmental
Protection Agency. March 31.
333 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. EPA-HQ-OAR-2009-0472-0243
334 Neumann, I.E., M.T. Dickie, and R.E. Unsworth. (1994). Linkage Between Health Effects
Estimation and Morbidity Valuation in the Section 812 Analysis—Draft Valuation Document.
Industrial Economics Incorporated (lEc) Memorandum to Jim DeMocker, U.S.
Environmental Protection Agency, Office of Air and Radiation, Office of Policy Analysis and
Review. March 31.
335 Tolley, G.S. et al. January (1986). Valuation of Reductions in Human Health Symptoms
and Risks. University of Chicago. Final Report for the U.S. Environmental Protection
Agency.
336 Council of Economic Advisors. (2005). The Annual Report of the Council of Economic
Advisors. In: Economic Report of the President. Table B-60. U.S. Government Printing
Office: Washington, DC.
337 National Research Council (NRC). (2002). Estimating the Public Health Benefits of
Proposed Air Pollution Regulations. The National Academies Press: Washington, D.C.
338 U.S. Environmental Protection Agency, (2004a). Final Regulatory Analysis: Control of
Emissions from Nonroad Diesel Engines. EPA420-R-04-007. Prepared by Office of Air and
Radiation. Retrieved on April 10, 2009, from http://www.epa.gov/nonroad-
diesel/2004fr/420r04007.pdf. EPA-HQ-OAR-2009-0472-0140
339 U.S. Environmental Protection Agency, (2005). Regulatory Impact Analysis for the Clean
Air Interstate Rule. EPA 452/-03-001. Prepared by Office of Air and Radiation. Retrieved on
April 10, 2009, from http://www.epa.gov/interstateairquality/tsd0175.pdf.
340 U.S. Environmental Protection Agency, (2006). Regulatory Impact Analysis for the PM
NAAQS. EPA Prepared by Office of Air and Radiation. Retrieved on April 10, 2009, from
http://www.epa.gov/ttn/ecas/regdata/RIAs/Chapter%205-Benefits.pdf. EPA-HQ-OAR-2009-
0472-0240
7-179
-------
Regulatory Impact Analysis
341 Industrial Economics, Inc. (2006). Expanded Expert Judgment Assessment of the
Concentration-Response Relationship Between PM2.5 Exposure and Mortality. Prepared for
EPA Office of Air Quality Planning and Standards, September. Retrieved on April 10, 2009,
from http://www.epa.gov/ttn/ecas/regdata/Uncertainty/pm ee report.pdf. EPA-HQ-OAR-
2009-0472-0242
342 National Research Council (NRC), 2008. Estimating Mortality Risk Reduction and
Economic Benefits from Controlling Ozone Air Pollution. The National Academies Press:
Washington, D.C. EPA-HQ-OAR-2009-0472-0322
343 National Research Council (NRC). 2002. Estimating the Public Health Benefits of
Proposed Air Pollution Regulations. The National Academies Press: Washington, D.C.
344 U.S. Environmental Protection Agency. October 2006. Final Regulatory Impact Analysis
(RIA)for the Proposed National Ambient Air Quality Standards for Paniculate Matter.
Prepared by: Office of Air and Radiation. Available at http://www.epa.gov/ttn/ecas/ria.html.
EPA-HQ-OAR-2009-0472-0240
345 U.S. Environmental Protection Agency (U.S. EPA). 2008. Regulatory Impact Analysis,
2008 National Ambient Air Quality Standards for Ground-level Ozone, Chapter 6. Office of
Air Quality Planning and Standards, Research Triangle Park, NC. March. Available at
. EPA-HQ-OAR-2009-
0472-0238
346 U.S. Environmental Protection Agency (U.S. EPA). 2009. Regulatory Impact Analysis:
National Emission Standards for Hazardous Air Pollutants from the Portland Cement
Manufacturing Industry. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. April. Available on the Internet at
. EPA-HQ-OAR-
2009-0472-0241
347 U.S. Environmental Protection Agency (U.S. EPA). 2010. Final NO2 NAAQS Regulatory
Impact Analysis (RIA). Office of Air Quality Planning and Standards, Research Triangle
Park, NC. April. Available on the Internet at
http://www.epa.gov/ttn/ecas/regdata/RIAs/FinalNO2RIAfulldocument.pdf. Accessed March
15, 2010. EPA-HQ-OAR-2009-0472-0237
348 U.S. Environmental Protection Agency (U.S. EPA). 2008. 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 .
349 Fann, N. et al. (2009). The influence of location, source, and emission type in estimates of
the human health benefits of reducing a ton of air pollution. Air Qual Atmos Health.
Published online: 09 June, 2009. EPA-HQ-OAR-2009-0472-0229
7-180
-------
Environmental and Health Impacts
350 Pope, C.A., III, R.T. Burnett, MJ. Thun, E.E. Calle, D. Krewski, K. Ito, and G.D.
Thurston. 2002. "Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine
Particulate Air Pollution." Journal of the American Medical Association 287:1132-1141.
EPA-HQ-OAR-2009-0472-0263
351 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. EPA-HQ-OAR-2009-0472-
1661
352 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.
Techno!., 42(7):2268-2274. EPA-HQ-OAR-2009-0472-0242
353 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 http://www.epa.gov/ttn/ecas/regdata/Uncertaintv/pm_ee_report.pdf EPA-HQ-
OAR-2009-0472-0242
354 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. EPA-HQ-OAR-2009-
0472-1677
355 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(l):5-76. EPA-
HQ-OAR-2009-0472-0245
356 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. EPA-HQ-OAR-2009-0472-0235
357 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
. EPA-
HQ-OAR-2009-0472-0232
358IPCC WGI, 2007. The baseline temperature increases by 2100 from our MiniCAM-
MAGICC runs are 1.8°C to 4.5°C.
359 IPCC. 2007. Climate Change 2007 - Synthesis Report Contribution of Working Groups I,
II and III to the Fourth Assessment Report of the IPCC.
7-181
-------
Regulatory Impact Analysis
360 Interagency Working Group on Social Cost of Carbon, U.S. Government, with
participation by Council of Economic Advisers, Council on Environmental Quality,
Department of Agriculture, Department of Commerce, Department of Energy, Department of
Transportation, Environmental Protection Agency, National Economic Council, Office of
Energy and Climate Change, Office of Management and Budget, Office of Science and
Technology Policy, and Department of Treasury, "Social Cost of Carbon for Regulatory
Impact Analysis Under Executive Order 12866," February 2010, available in docket EPA-
HQ-OAR-2009-0472.
361 "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
362 "2008/9 Blueprint for Sustainability - Our Future Works" Ford Motor Company. Last
accessed on the Internet on March 24, 2010 at the following URL:
http://www.ford.com/go/sustainability and available in Docket EPA-HQ-OAR-2009-0472.
363 "Japan's new diet plan: Slim down, ounce by ounce" Automotive News, July 7, 2008.
364 "Mazda: Don't believe the hot air being emitted by hybrid hype" Automotive News,
March 30, 2009.
365 "Improved aerodynamics, lightweight construction, tyres and drive technology -
BlueEFFICIENCY in the C-Class: Fuel consumption reduced by 12 per cent" Daimler Press
Release, February 28, 2008. Last accessed on the Internet on March 24, 2010 at the following
URL: http://media.daimler.com/dcmedia/0-921-1050602-l-1050604-l-0-0-0-0-l-11701-
854934-0-l-0-0-0-0-0.html and available in Docket EPA-HQ-OAR-2009-0472.
366 "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.
367 "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
368 "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
369 "Supplemental Results on the Independent Effects of Curb Weight, Wheelbase and Track
on Fatality Risk", Dynamic Research, Inc., DRI-TR-05-01, May 2005
370 "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.
371 "Blood and Oil: Vehicle Characteristics in Relation to Fatality Risk and Fuel Economy",
L.S. Robertson, American Journal of Public Health, Vol. 96, No. 11, November 2006.
7-182
-------
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.5 of the preamble to the rale 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 rale 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 compares two effects. On the one hand, the vehicles will become more
expensive, which would, by itself, discourage sales. On the other hand, the vehicles will have
improved fuel economy and thus lower operating costs, which makes them more attractive to
consumers. If consumers do not accurately compare the value of fuel savings with the
increased cost of fuel economy technology in their vehicle purchase decisions, as discussed in
Preamble III.H. 1, they will continue to behave in this way after this rale. If auto makers have
accurately gauged how consumers consider fuel economy when purchasing vehicles and have
provided the amount that consumers want in vehicles, then consumers should not be expected
to want the more fuel-efficient vehicles. After all, auto makers would have provided as much
fuel economy as consumers want. If, on the other hand, auto makers underestimated
consumer demand for fuel economy, as suggested by some commenters and discussed in
Preamble Section III.H. 1 and RIA Section 8.1.2, then this rule may lead to production of more
desirable vehicles, and vehicle sales may increase. This assumption implies that auto makers
have missed some profit-making opportunities. The results presented in this analysis depend
on the assumption that more fuel efficient vehicles that yield net consumer benefits over five
years would not otherwise be offered on the vehicle market due to market failures on the part
of vehicle manufacturers. If vehicles that achieve the fuel economy standards prescribed by
today's ralemaking would already be available, but consumers chose not to purchase them,
then this ralemaking would not result in an increase in vehicle sales, because it does not alter
how consumers make decisions about which vehicles to purchase
The analysis starts with the increase in costs estimated by the OMEGA model. 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
8-1
-------
Regulatory Impact Analysis
financing costs, as well as increases in the resale value of the vehicle. These factors weigh
against each other: the higher sales tax, insurance, and financing 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-
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 February 9, 2010, the national average interest rate for a 5 year new car loan was
6.54 percent.6 Converting the up-front payment to an annual value paid over five years results
in a consumer paying 24.1% of the up-front amount every year. The present value of these
five payments results in an increase of 10.3% of the cost, using a 3% discount rate; with a 7%
discount rate, the increase is -1.2%. NHTSA's RIA notes that 70% of auto purchases use
financing; applying that fraction to this cost increase results in an addition of 7.2% in
financing costs with a 3% discount rate, and -0.9% 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 + 0.072 - 0.302) = 0.914 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 - 0.009 - 0.250) =
0.876 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 6.3 of this
8-2
-------
Other Economic and Social Impacts
RIA 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 per year 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 -I8: 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
the demand elasticity for new vehicles under this rule is the same as the elasticity for new
vehicles in the past. 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
67,500
76,000
114,000
222,200
360,500
% CHANGE
0.7
0.8
1.1
2.1
3.3
CHANGE IN
TRUCK SALES
62,100
190,200
254,900
352,800
488,000
% CHANGE
1.1
3.2
4.3
6.1
8.6
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
8-3
-------
Regulatory Impact Analysis
those for cars (except in 2012). This approach may not accurately reflect the role of fuel
savings in consumers' purchase decisions, as the discussion in Preamble Section III.H.l
suggests. If consumers consider fuel savings in a different fashion than modeled here, then
this approach will not accurately reflect the impact of this rule on vehicle sales.
Table 8-2 Vehicle Sales Impacts Using a 7% Discount Rate
2012
2013
2014
2015
2016
CHANGE IN
CAR SALES
62,800
70,500
106,100
208,400
339,400
% CHANGE
0.7
0.7
1
2
3.1
CHANGE IN
TRUCK SALES
58,300
92,300
127,700
194,200
280,000
% CHANGE
1
1.5
2.1
3.3
4.9
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. Some of the factors that might affect consumer vehicle purchases
include changing market conditions, changes in vehicle characteristics that might accompany
improvements in fuel economy, or consumers considering a different "payback period" for
their fuel economy purchases. These complications add considerable uncertainty to our
vehicle sales impact analysis.
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 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
8-4
-------
Other Economic and Social Impacts
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 rule.
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
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
8-5
-------
Regulatory Impact Analysis
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.1
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%.
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
1 Q
decisions, use aggregated data that can cover long time periods.
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.
8-6
-------
Other Economic and Social Impacts
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.
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
8-7
-------
Regulatory Impact Analysis
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: Recent research39 presents 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 per year 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.
A new review from David Greene41 suggests continued lack of convergence on the
value of increased fuel economy to consumers. Of 27 studies, willingness to pay for fuel
economy as a percent of the expected value of fuel savings varied from highly positive to
highly negative. Significant numbers of studies found that consumers overvalued fuel
economy, undervalued fuel economy, or roughly valued fuel economy correctly relative to
fuel savings. 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 a result, "fuel economy" may
in fact represent several characteristics at the same time. Indeed, Gramlich42 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. Greene finds that, while some of the
variation may be explainable due to issues in some of the studies, the variation shows up in
studies that appear to be well conducted. As a result, further work needs to be conducted
before it is possible to identify one value that represents the role of fuel economy in consumer
purchase decisions.
Some studies43 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 studies44 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 studies45 notes that its baseline model implies that
8-8
-------
Other Economic and Social Impacts
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? While existing research does not offer full
answers, it is important to attempt to explore these questions, because under certain
assumptions, the purely private benefits of fuel economy (fuel savings, time savings, increases
in driving time) are likely to be accompanied by private losses. If there is no such offset, or if
it is small or insignificant, the reason lies in some kind of market failure.
A detailed literature attempts to identify possible market failures that would justify the
assumption that the degree of consumer welfare loss is relatively small. 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,46 since consumers appear to undervalue a wide range of investments in energy
conservation. Some possible explanations for the paradox47 include:
• Consumers put little weight on benefits from fuel economy in the future and show
high discount rates;
• Consumers do not find the benefits from fuel economy to be sufficiently salient at
the time of purchase, even if it would be in consumers' economic interest to take
account of those benefits;
• Consumers consider other attributes more important than fuel economy at the time
of vehicle purchase, especially if fuel economy is a relatively "shrouded" attribute;
• 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, while at the time of purchase the increased
costs of fuel-saving technologies are certain and immediate;
• Consumers may not be able to find the vehicles they want with improved fuel
economy;
8-9
-------
Regulatory Impact Analysis
• 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.
Both theoretical and empirical research suggests that, in the context of fuel economy
and elsewhere, many consumers do not make energy-efficient investments even when those
investments would pay off in the relatively short-term.48 This conclusion is in line with
related findings that consumers may underweight benefits and costs that are less salient or that
will be realized only in the future.4 At the same time, it is worth noting that many of these
behaviors can be accounted for in standard economic models. For instance, accounting for
uncertainty in future fuel savings is a common practice. For some consumers, high observed
discount rates may reflect the illiquidity of investments in fuel economy or high opportunity
costs of such investments where consumers are carrying high-interest-rate debt. There is
disagreement in the literature about the degree to which these explanations contribute to
understanding the Energy Paradox, and additional empirical investigation is still needed.
The producer side of this paradox is much less studied. Hypotheses for
underprovision of fuel economy by producers, related to those involving consumers, include:
• Producers put more effort into attributes that consumers have regularly sought in
the past, such as size and power, than into fuel and time 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 and time savings 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 underestimated 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 savings from fuel economy improvements. 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, helping to answer some of the central questions about
8-10
-------
Other Economic and Social Impacts
relevant effects on consumer welfare. 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.
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,50 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.51
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.52 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 complication associated with consumer choice modeling is accurate
prediction of producers' responses to the rule. Auto makers not only predict consumers'
preferences for vehicles; they also may seek to influence those preferences through marketing
and advertising.53 In addition, auto makers are commonly considered to have market power;
they can influence the prices that consumers pay to increase their profits. As a result, the
price increases that consumers face may reflect strategic factors that could make them higher
or lower than the technology costs. Including these market features into consumer vehicle
choice models is a complex undertaking. Not all consumer vehicle choice models include a
producer model, and those that do may not include much detail, due to computational limits.
Technology costs still represent an accurate measure of the opportunity cost of resources to
society, but they may overestimate or underestimate the effect on the prices that consumers
8-11
-------
Regulatory Impact Analysis
face. Firms with market power usually pass along less than full cost increases, in order not to
reduce sales very much. As a result, for most vehicles the increased technology costs may not
equal the price increases that consumers will see.
An additional feature of consumer choice models, as noted above, is that they can be
used to calculate consumer surplus impacts on vehicle purchase decisions. 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.
Principles of welfare analysis can be useful for understanding the role of consumer
vehicle choice models in benefit-cost analysis. Consumer welfare is commonly measured as
the change in income that would leave the consumer as well off in the presence of the change
(in this case, the increase in fuel economy and vehicle price) as in the absence of the change;
this amount is known as compensating variation, since the consumer is compensated for the
change.A If the vehicle has not changed other than those two characteristics, then a consumer
has the choices of (i) paying the higher price for the vehicle, (ii) choosing to buy a different
vehicle, or (iii) not buying a new vehicle. The only reason the consumer would decide to pay
the higher price, (i), is if this option is preferable to the other two options. If the consumer
cares nothing about the increased fuel economy but is given an amount of money equal to the
price increase, she is at least as well off as before the price increase: she can still buy the
original vehicle (which has improved fuel economy but is otherwise identical), or she can still
choose options (ii) or (iii). Thus, the price increase due to the rule is an upper bound on the
consumer's welfare loss, if no other vehicle characteristics of interest to the consumer has
changed. If the consumer actually appreciates the improved fuel economy, the welfare loss is
even smaller. However, if the vehicle has changed due to the fuel economy increase in ways
other than price and fuel economy, or if there are additional costs associated with these
vehicles not included in the analysis, then there may be additional welfare impacts that are not
included in the technology cost estimates.
At this point, it is unclear whether two consumer vehicle choice models given the
same scenario would produce similar results in either prediction of changes in the vehicles
A A closely related concept, equivalent variation, measures the change in income that would be a perfect
substitute for the increased fuel economy and vehicle cost. These measures differ based on whether the basis for
evaluation is the consumer's utility before the change (compensating variation) or the consumer's utility after the
change (equivalent variation). In practice, the difference between these two measures is typically very small for
marketed goods.
8-12
-------
Other Economic and Social Impacts
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 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.
Given the current limitations in modeling the role of fuel economy in vehicle purchase
decisions, and limitations in modeling market responses to the new regulations, in this
rulemaking EPA compares the fuel and other savings that consumers will receive with the
technology costs of the vehicles. The regulations have been carefully designed so that the full
range of vehicle choices in the marketplace could be maintained; rigorous technological
feasibility, cost, and lead-time analysis has shown that the standards could be met while
maintaining current levels of other vehicle attributes. For these reasons, EPA believes that
consumers will enjoy significant savings that substantially outweigh any likely consumer
welfare losses. Nonetheless, EPA continues to consider these questions, and 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, the estimates of consumers' willingness to pay for increased fuel economy,
and overall efects on consumer welfare.
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 future
rules.
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 these standards. In other words, how long would it take for the expected fuel
savings to outweigh the increased cost of a new vehicle? For example, a new 2016 MY
vehicle is estimated to cost $948 more (on average, and relative to the reference case vehicle)
due to the addition of new GHG reducing technology (see 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 6 for details on fuel savings). But how many months or
years would pass before the fuel savings exceed the upfront cost of $948?
Table 8-3 provides the answer to this question for a vehicle purchaser who pays for the
new vehicle upfront in cash (we discuss later in this section the payback period for consumers
8-13
-------
Regulatory Impact Analysis
who finance the new vehicle purchase with a loan). The table uses annual miles driven
(vehicle miles traveled, or VMT) and survival rates consistent with the emission and benefits
analyses presented in Chapter 4 of the joint TSD. We have included rebound VMT in the
control case but not in the reference case, consistent with other parts of our analysis. We have
also included fuel savings associated with A/C controls (in the control case only), but have
not included expected A/C-related maintenance savings. We discuss the likely maintenance
savings in Chapter 2 of this RIA. Further, this analysis does not include other societal
impacts such as the value of increased driving, or noise, congestion and accidents since we
really want to focus on those factors consumers consider most while in the showroom
considering a new car purchase. Car/truck fleet weighting is handled as described in Chapter
1 of the joint TSD. As can be seen in the table, it will take under 3 years (2 years and 7
months at a 3% discount rate, 2 years and 9 months at a 7% discount rate) for the cumulative
fuel savings to exceed the upfront increase in vehicle cost. For the average driver, this
payback would occur at around 46,000 to 49,000 miles, depending on the discount rate. For
the driver that drives more than the average, the payback would come sooner. For the driver
that drives less than the average, the payback would come later.
Table 8-3 Payback Period on a 2016MY New Vehicle Purchase via Cash (2007 dollars)
Year of
Ownership
1
2
3
4
Increased
Vehicle
Costa
($)
$1,018
Fuel Priceb
($/gal)
$3.07
$3.13
$3.19
$3.22
Reference
VMTC
(miles)
17,850
17,297
16,789
16,133
Control
VMTC
(miles)
18,186
17,623
17,105
16,437
Reference
Fuel
Costs'1
($)
$2,437
$2,410
$2,377
$2,310
Control
Fuel
Costs'1
($)
$2,013
$1,990
$1,963
$1,908
Annual
Fuel
Savings
($)
$424
$420
$414
$402
Cumulative
Discounted
Fuel
Savings at
3%
($)
$418
$820
$1,204
$1,567
Cumulative
Discounted
Fuel
Savings at
7%
($)
$410
$790
$1,139
$1,457
a Increased cost of the rule is $948; the value here includes nationwide average sales tax of 5.3% and increased
insurance premiums of 1.98%; both of these percentages are discussed in section 8.1.1.
b AEO 2010 Early Release reference case fuel price including taxes.
c VMT is calculated as the weighted car/truck VMT with cars estimated to account for 66% of the fleet and
trucks 34%; 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 per 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 February 9, 2010, the national
average interest rate for a 5 year new car loan was 6.54 percent. If the increased vehicle cost
is spread out over 5 years at 6.54 percent, the analysis would look like that shown in Table
8-4. As can be seen in this table, the fuel savings immediately outweigh the increased
payments on the car loan, amounting to $177 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
8-14
-------
Other Economic and Social Impacts
new vehicle their monthly fuel savings would be greater than the increase in the loan
payments. This amounts to a savings on the order of $9 to $15 per month throughout the
duration of the 5 year loan. Note that in year six when the car loan is paid off, the net
savings equal the fuel savings (as would be the case for the remaining years of ownership).
Table 8-4 Payback Period on a 2016 MY New Vehicle Purchase via Credit (2007 dollars)
Year of
Ownership
1
2
3
4
5
6
Increased
Vehicle
Costa
($)
$245
$245
$245
$245
$245
$0
Fuel
Priceb
($/gal)
$3.07
$3.13
$3.19
$3.22
$3.27
$3.29
Reference
VMTC
(miles)
17,850
17,297
16,789
16,133
15,451
14,668
Control
VMTC
(miles)
18,186
17,623
17,105
16,437
15,742
14,944
Reference
Fuel
Costs'1
($)
$2,437
$2,410
$2,377
$2,310
$2,244
$2,148
Control
Fuel
Costs'1
($)
$2,013
$1,990
$1,963
$1,908
$1,853
$1,774
Annual
Fuel
Savings
($)
$424
$420
$414
$402
$391
$374
Annual
Discounted
Net
Savings at
3%
($)
$177
$167
$157
$142
$127
$318
Annual
Discounted
Net
Savings at
7%
($)
$173
$158
$142
$124
$107
$258
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 6.54
percent.
b AEO 2010 Early Release reference case fuel price including taxes.
c VMT is calculated as the weighted car/truck VMT with cars estimated to account for 66% of the fleet and
trucks 34%; 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 per 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 $3,100 at a 3%
discount rate, or $2,300 at a 7% discount rate.
Table 8-5 Lifetime Discounted Net Savings on a 2016 MY New Vehicle Purchase (2007 dollars)
Purchase Option
Increased
Discounted Vehicle
Cost
($)
Lifetime
Discounted Fuel
Savings'5'0
($)
Lifetime
Discounted Net
Savings
($)
3% discount rate
Cash
Credit3
$1,018
$1,140
$4,306
$4,306
$3,303
$3,166
7% discount rate
Cash
Credit2
$1,018
$1,040
$3,381
$3,381
$2,396
$2,340
1 Assumes a 5 year loan at 6.54 percent.
8-15
-------
Regulatory Impact Analysis
VMT is calculated as the weighted car/truck VMT with cars estimated to account for
66% of the fleet and trucks 34%; VMT shown here includes survival fraction and, for
the control case, rebound VMT.
°Fuel savings here were calculated using AEO 2010 Early Release reference case fuel
price including taxes.
8.2 Energy Security Impacts
This chapter will only describe the energy security analysis that was conducted
beyond that described in Chapter 4 of the TSD. Additional analysis was conducted to provide
inputs to EPA's OMEGA model. For a detailed discussion of the development of the energy
security estimates, please refer to Chapter 4 of the joint TSD.
After the EPA-sponsored peer review of the Oak Ridge National Laboratory's
(ORNL) Energy Security Analysis was completed in 2008, ORNL, at EPA's request, updated
the analysis using values from the AEO 2009 rather than the 2007 values. The methodology
used to update this analysis was the same one that was peer-reviewed.54 The results are
shown in Table 8-6. ORNL estimated the energy security premium for 2015, 2020, and 2030.
Since the AEO 2009 forecasts ends in 2030, EPA assumed that the post-2030 energy security
premium did not change through 2040.
Table 8-6 Energy Security Premium in 2015,2020,2030, and 2040 (2007$/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 -$27.89)
EPA linearly interpolated the values for the years 2016 through 2019, using the 2015
and 2020 values as endpoints. EPA followed the same procedure to estimate the 2021
through 2029 estimates, using the 2020 and 2030 values as endpoints. Post-2030, EPA
assumed that the energy security estimate did not change. The final set of values that was
used by the OMEGA model is shown in Table 8-7.
8-16
-------
Other Economic and Social Impacts
Table 8-7 Energy Security Premium Estimates for Years 2015-2040 (2007$/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 RIA. EPA used the same assumption that NHTSA used in its
Corporate Average Fuel Economy and CAFE Reform for MY 2008-2011 Light Trucks rule,
which assumed each gallon of fuel saved reduces total U.S. imports of crude oil or refined
,B56
products by 0.95 gallons . Section 5.3 of this RIA contains a discussion regarding caveats
B 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
8-17
-------
Regulatory Impact Analysis
for the fuel savings estimated due to implementation of this rule. Section III.H. 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.
Table 8-8 Total Annual Energy Security Benefits in 2015,2020,2030, and 2040 (Billions of 2007 dollars)
YEAR
2015
2020
2030
2040
BENEFITS
$0.57
$2.17
$4.55
$6.00
8.3 Other Impacts
There are other impacts associated with the 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 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 these
standards, but they are nevertheless important to include. We summarize the value of these
other impacts in section 8.4.4 of this RIA. Please refer to the joint TSD for more information
about these impacts and how EPA and NHTSA use them in their analyses.
8.3.1 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
Economic Growth, and Low Economic Growth Scenarios presented in the Energy Information Administration's
Annual Energy Outlook 2007, NHTSA estimated that approximately 50 percent of the reduction in fuel
consumption is likely to be reflected in reduced U.S. imports of refined fuel, while the remaining 50 percent
would be expected to be reflected in reduced domestic fuel refining. Of this latter figure, 90 percent is
anticipated to reduce U.S. imports of crude petroleum for use as a refinery feedstock, while the remaining 10
percent is expected to reduce U.S. domestic production of crude petroleum. Thus on balance, each gallon of fuel
saved is anticipated to reduce total U.S. imports of crude petroleum or refined fuel by 0.95 gallons.
8-18
-------
Other Economic and Social Impacts
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
8.3.2 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 joint TSD. Again, the
negative effect that rebound driving has on the fuel consumption savings associated with the
GHG standards is included in the fuel economy savings presented in section 8.5 of this RIA.
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.
8.3.3 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
8-19
-------
Regulatory Impact Analysis
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).
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.
8-20
-------
Other Economic and Social Impacts
Table 8-9 $/mile Inputs used for External Costs
EXTERNAL COSTS
Congestion
Accidents
Noise
$/VMT
$
$
$
0.052
0.023
0.001
8.3.4 Summary of Other Impacts
Table 8-10 summarizes the other economic impacts discussed in sections 8.3.1
through 8.3.3.
8-21
-------
Regulatory Impact Analysis
Table 8-10 Other Impacts Associated with the 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%
VALUE OF
REDUCED
REFUELING
$100
$300
$500
$700
$1,100
$1,500
$1,800
$2,100
$2,400
$2,700
$3,000
$3,300
$3,600
$3,800
$4,000
$4,200
$4,400
$4,600
$4,800
$4,900
$5,100
$5,300
$5,400
$5,600
$5,700
$5,900
$6,000
$6,200
$6,300
$6,500
$6,600
$6,800
$7,000
$7,100
$7,300
$7,500
$7,700
$7,800
$8,000
$87,900
$40,100
VALUE OF
INCREASED
DRIVING
$200
$400
$700
$1,200
$1,800
$2,400
$3,000
$3,600
$4,200
$4,700
$5,300
$5,800
$6,200
$6,700
$7,200
$7,600
$8,100
$8,500
$8,800
$9,200
$9,600
$10,000
$10,400
$10,900
$11,300
$11,700
$12,100
$12,500
$13,000
$13,500
$13,900
$14,400
$14,900
$15,500
$16,000
$16,600
$17,200
$17,800
$18,400
$171,500
$75,500
ACCIDENTS,
NOISE,
CONGESTION
-$100
-$200
-$400
-$700
-$1,000
-$1,400
-$1,700
-$2,000
-$2,300
-$2,600
-$2,900
-$3,200
-$3,400
-$3,700
-$3,900
-$4,100
-$4,300
-$4,500
-$4,600
-$4,800
-$4,900
-$5,100
-$5,200
-$5,400
-$5,500
-$5,700
-$5,800
-$5,900
-$6,100
-$6,300
-$6,400
-$6,600
-$6,700
-$6,900
-$7,100
-$7,200
-$7,400
-$7,600
-$7,800
-$84,800
-$38,600
8-22
-------
Other Economic and Social Impacts
8.4 Summary of Costs and Benefits
In this section we present a summary of costs, benefits, and net benefits of the rule.
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 3 and 7 percent discount rates. In this table, fuel savings are
calculated using pre-tax fuel prices.
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
$15,600
-$35,700
-$20,100
2030
$15,800
-$79,800
-$64,000
2040
$17,400
-$119,300
-$101,900
2050
$19,000
-$171,200
-$152,200
NPV, 3%
$345,900
-$1,545,600
-$1,199,700
NPV, 7%
$191,900
-$672,600
-$480,700
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 four SCC values considered by EPA. As discussed in [the SCC TSD for this final
rule], the models used to estimate SCC may not capture the economic effects of all possible
adverse consequences of climate change and may therefore lead to underestimates of the
SCC.
In addition the monetized GHG benefits presented below exclude the value of
reductions in non-CCh GHG emissions (HFC, CH4, NiO) expected under this final rule.
Although EPA has not monetized the benefits of reductions in non-COi GHGs, the value of
these reductions should not be interpreted as zero. Rather, the reductions in non-CCh GHGs
will contribute to this rule's climate benefits, as explained in Section III.F.2. The technical
support document, Social Cost of Carbon for Regulatory Impact Analysis Under Executive
Order 12866, (i.e., SCC TSD) notes the difference between the social cost of non-CO2
emissions and CO2 emissions, and specifies a goal to develop methods to value non-CCh
emissions in future analyses.64
8-23
-------
Regulatory Impact Analysis
Table 8-12 Use Estimated Societal Benefits Associated with the Light-Duty Vehicle GHG Program
(Millions of 2007 dollars)
BENEFITS
2020
2030
Reduced CO2 Emissions at each assumed SCC value"'0
Avg SCC at 5% $900 $2,700
Avg SCC at 3% $3,700 $8,900
Avg SCC at 2.5% $5,800 $14,000
95th percentile SCC at 3% $1 1,000 $27,000
Criteria Pollutant
Benefitsd'e'f'8
Energy Security Impacts
(price shock)
Reduced Refueling
Value of Increased Driving11
Accidents, Noise,
Congestion
Quantified Annual Benefits at
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC at 3%
B
$2,200
$2,400
$4,200
-$2,300
each assumed
$7,400
$10,200
$12,300
$17,500
$1,200-
$1,300
$4,500
$4,800
$8,800
-$4,600
SCC valueb'c
$17,500
$23,700
$28,800
$41,800
2040
$4,600
$14,000
$21,000
$43,000
$1,200-
$1,300
$6,000
$6,300
$13,000
-$6,100
$25,100
$34,500
$41,500
$63,500
2050
$7,200
$21,000
$30,000
$62,000
$1,200-
$1,300
$7,600
$8,000
$18,400
-$7,800
$34,700
$48,500
$57,500
$89,500
NPV, 3%A
$34,500
$176,700
$299,600
$538,500
$21,000
$81,900
$87,900
$171,500
-$84,800
$312,000
$454,200
$577,100
$816,000
NPV, 7%A
$34,500
$176,700
$299,600
$538,500
$14,000
$36,900
$40,100
$75,500
-$38,600
$162,400
$304,600
$427,500
$666,400
aNote that net present value of reduced CO2 emissions is calculated differently than other benefits. The same
discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is used to
calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
b Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O)
expected under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2
emissions, the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2
GHGs will contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the
difference between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop
methods to value non-CO2 emissions in future analyses.
c Section 7.5 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates
range as follows: for Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $35-$65; and for 95th percentile SCC at 3%: $65-$136. Section 7.5 also presents these SCC estimates.
dNote that "B" indicates unquantified criteria pollutant benefits in the year 2020. For the final rule, we only
modeled the rule's PM2.5- and ozone-related impacts in the calendar year 2030. For the purposes of estimating a
stream of future-year criteria pollutant benefits, we assume that the benefits out to 2050 are equal to, and no less
than, those modeled in 2030 as reflected by the stream of estimated future emission reductions. The NPV of
criteria pollutant-related benefits should therefore be considered a conservative estimate of the potential benefits
associated with the final rule.
e The benefits presented in this table include an estimate of PM-related premature mortality derived from Laden
et al., 2006, and the ozone-related premature mortality estimate derived from Bell et al., 2004. If the benefit
estimates were based on the ACS study of PM-related premature mortality (Pope et al., 2002) and the Levy et al.,
2005 study of ozone-related premature mortality, the values would be as much as 70% smaller.
f The calendar year benefits presented in this table assume either a 3% discount rate in the valuation of PM-
related premature mortality ($1,300 million) or a 7% discount rate ($1,200 million) to account for a twenty-year
segmented cessation lag. Note that the benefits estimated using a 3% discount rate were used to calculate the
NPV using a 3% discount rate and the benefits estimated using a 7% discount rate were used to calculate the
NPV using a 7% discount rate. For benefits totals presented at each calendar year, we used the mid-point of the
criteria pollutant benefits range ($1,250).
8 Note that the co-pollutant impacts presented here do not include the full complement of endpoints that, if
quantified and monetized, would change the total monetized estimate of impacts. The full complement of
8-24
-------
Other Economic and Social Impacts
human health and welfare effects associated with PM and ozone remain unquantified because of current
limitations in methods or available data. We have not quantified a number of known or suspected health effects
linked with ozone and PM for which appropriate health impact functions are not available or which do not
provide easily interpretable outcomes (e.g., changes in heart rate variability). Additionally, we are unable to
quantify a number of known welfare effects, including reduced acid and particulate deposition damage to
cultural monuments and other materials, and environmental benefits due to reductions of impacts of
eutrophication in coastal areas.
h Calculated using pre-tax fuel prices.
Table 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-
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 four SCC
values considered by EPA.
Table 8-13 Quantified Net Benefits Associated with the Light-Duty Vehicle GHG Program3
(Millions of 2007 dollars)
Quantified Annual
Costs
2020
-$20,100
2030
-$64,000
2040
-$101,900
2050
-$152,200
NPV, 3%
-$1,199,700
NPV, 7%
-$480,700
Quantified Annual Benefits at each assumed SCC value"'0
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC
at 3%
$7,400
$10,200
$12,300
$17,500
$17,500
$23,700
$28,800
$41,800
$25,100
$34,500
$41,500
$63,500
$34,700
$48,500
$57,500
$89,500
$312,000
$454,200
$577,100
$816,000
$162,400
$304,600
$427,500
$666,400
Quantified Net Benefits at each assumed SCC value"'0
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC
at 3%
$27,500
$30,300
$32,400
$37,600
$81,500
$87,700
$92,800
$105,800
$127,000
$136,400
$143,400
$165,400
$186,900
$200,700
$209,700
$241,700
$1,511,700
$1,653,900
$1,776,800
$2,015,700
$643,100
$785,300
$908,200
$1,147,100
aFuel impacts were calculated using pre-tax fuel prices.
b Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O)
expected under this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2
emissions, the value of these reductions should not be interpreted as zero. Rather, the reductions in non-CO2
GHGs will contribute to this rule's climate benefits, as explained in Section III.F.2. The SCC TSD notes the
difference between the social cost of non-CO2 emissions and CO2 emissions, and specifies a goal to develop
methods to value non-CO2 emissions in future analyses.
0 Section 7.5 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates
range as follows: for Average SCC at 5%: $5-$16; for Average SCC at 3%: $21-$45; for Average SCC at
2.5%: $35-$65; and for 95th percentile SCC at 3%: $65-$136. Section 7.5 also presents these SCC estimates.
Note also that net present value of reduced CO2 emissions is calculated differently than other benefits. The
same discount rate used to discount the value of damages from future emissions (SCC at 5, 3, 2.5 percent) is
used to calculate net present value of SCC for internal consistency. Refer to the SCC TSD for more detail.
EPA also conducted a separate analysis of the total benefits over the model year
lifetimes of the 2012 through 2016 model year vehicles. In contrast to the calendar year
analysis presented in Table 8-11 through Table 8-13, the model year lifetime analysis shows
the lifetime impacts of the program on each of these MY fleets over the course of its lifetime.
Full details of the inputs to this analysis can be found in Chapter 5 of this RIA. The societal
8-25
-------
Regulatory Impact Analysis
benefits of the full life of each of the five model years from 2012 through 2016 are shown in
Table 8-14 and Table 8-15 at both a 3 percent and a 7 percent discount rate, respectively. The
net benefits are shown in Table 8-16 and Table 8-17 for both a 3 percent and a 7 percent
discount rate, respectively. Note that the quantified annual benefits shown in Table 8-14 and
Table 8-15 include fuel savings as a positive benefit. As such, the quantified annual costs as
shown in Table 8-16 and Table 8-17 do not include fuel savings since those are included as
benefits. Also note that Table 8-14 through Table 8-17 include the benefits of reduced CO2
emissions—and consequently the total benefits—for each of four SCC values considered by
EPA.
Table 8-14 Estimated Societal Benefits Associated with the Lifetimes of 2012-2016 Model year Vehicles
(Millions of 2007 dollars; 3% Discount Rate)
MONETIZED VALUES
Cost of Noise, Accident, Congestion ($)
Pretax Fuel Savings ($)
Energy Security ($) (price shock) a
Value of Reduced Refueling time ($)
Value of Additional Driving ($)
Value of PM2.5 related Health Impacts ($)b'c'd
Reduced CO2 Emissions at each assumed
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC at 3%
Total Benefits at each assumed SCC value e>t
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC at 3%
2012MY
-$1,100
$16,100
$900
$1,100
$2,400
$700
SCC value
$400
$1,700
$2,700
$5,100
$20,500
$21,800
$22,800
$25,200
2013MY
-$1,600
$23,900
$1,400
$1,600
$3,400
$900
$500
$2,400
$3,900
$7,300
$30,100
$32,000
$33,500
$36,900
2014MY
-$2,100
$32,200
$1,800
$2,100
$4,400
$1,300
$700
$3,100
$5,200
$9,600
$40,400
$42,800
$44,900
$49,300
2015MY
-$2,900
$46,000
$2,500
$3,000
$6,000
$1,800
$1,000
$4,400
$7,200
$13,000
$57,400
$60,800
$63,600
$69,400
2016MY
-$3,900
$63,500
$3,500
$4,000
$7,900
$2,400
$1,300
$5,900
$9,700
$18,000
$78,700
$83,300
$87,100
$95,400
SUM
-$11,600
$181,800
$10,100
$11,900
$24,000
$7,000
$3,800
$17,000
$29,000
$53,000
$227,000
$240,200
$252,200
$276,200
aNote that, due to a calculation error in the rule, the energy security impacts for the model year analysis were roughly half
what they should have been.
bNote that the co-pollutant impacts associated with the standards presented here do not include the full complement of
endpoints that, if quantified and monetized, would change the total monetized estimate of rule-related impacts. Instead, the
co-pollutant benefits are based on benefit-per-ton values that reflect only human health impacts associated with reductions in
PM2.5 exposure. Ideally, human health and environmental benefits would be based on changes in ambient PM2.5 and ozone
as determined by full-scale air quality modeling. However, EPA was unable to conduct a full-scale air quality modeling
analysis associated with the vehicle model year lifetimes for the final rule.
cThe PM2.5-related benefits (derived from benefit-per-ton values) presented in this table are based on an estimate of
premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six
Cities study (Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
d The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table assume a 3% discount rate in the
valuation of premature mortality to account for a twenty-year segmented cessation lag. If a 7% discount rate had been used,
the values would be approximately 9% lower.
e Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected under
this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions, the value of these
reductions should not be interpreted as zero. Rather, the reductions in non-CO2GHGs will contribute to this rule's climate
benefits, as explained in Section III.F.2. The SCC TSD notes the difference between the social cost of non-CO2 emissions
and CO2 emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses.
'Section 7.5 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates range as
follows: for Average SCC at 5%: $5-$l6; for Average SCC at 3%: $21-$45; for Average SCC at 2.5%: $35-$65; and for
95th percentile SCC at 3%: $65-$136. Section 7.5 also presents these SCC estimates. Note that net present value of reduced
CO2 emissions is calculated differently than other benefits. The same discount rate used to discount the value of damages
from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer
to the SCC TSD for more detail.
8-26
-------
Other Economic and Social Impacts
Table 8-15 Estimated Societal Benefits Associated with the Lifetimes of 2012-2016 Model year Vehicles
(Millions of 2007 dollars; 7% Discount Rate)
MONETIZED VALUES
Cost of Noise, Accident, Congestion ($)
Pretax Fuel Savings ($)
Energy Security ($) (price shock) a
Value of Reduced Refueling time ($)
Value of Additional Driving ($)
Value of PM2.5 related Health Impacts ($)b'c'd
Reduced CO2 Emissions at each assumed
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC at 3%
Total Benefits at each assumed SCC value e>1
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC at 3%
2012MY
-$900
$12,500
$800
$900
$1,900
$500
SCC value
$400
$1,700
$2,700
$5,100
$16,100
$17,400
$18,400
$20,800
2013MY
-$1,200
$18,600
$1,100
$1,300
$2,700
$800
e,f
$500
$2,400
$3,900
$7,300
$23,800
$25,700
$27,200
$30,600
2014MY
-$1,600
$25,100
$1,400
$1,700
$3,500
$1,000
$700
$3,100
$5,200
$9,600
$31,800
$34,200
$36,300
$40,700
2015MY
-$2,300
$36,000
$2,000
$2,400
$4,700
$1,400
$1,000
$4,400
$7,200
$13,000
$45,200
$48,600
$51,400
$57,200
2016MY
-$3,100
$49,600
$2,700
$3,200
$6,200
$1,900
$1,300
$5,900
$9,700
$18,000
$61,800
$66,400
$70,200
$78,500
SUM
-$9,200
$141,900
$8,000
$9,400
$19,000
$5,600
$3,800
$17,000
$29,000
$53,000
$178,500
$191,700
$203,700
$227,700
aNote that, due to a calculation error in the rule, the energy security impacts for the model year analysis were roughly half
what they should have been.
bNote that the co-pollutant impacts associated with the standards presented here do not include the full complement of
endpoints that, if quantified and monetized, would change the total monetized estimate of rule-related impacts. Instead, the
co-pollutant benefits are based on benefit-per-ton values that reflect only human health impacts associated with reductions in
PM2.5 exposure. Ideally, human health and environmental benefits would be based on changes in ambient PM2.5 and ozone
as determined by full-scale air quality modeling. However, EPA was unable to conduct a full-scale air quality modeling
analysis associated with the vehicle model year lifetimes for the final rule.
cThe PM2.5-related benefits (derived from benefit-per-ton values) presented in this table are based on an estimate of
premature mortality derived from the ACS study (Pope et al., 2002). If the benefit-per-ton estimates were based on the Six
Cities study (Laden et al., 2006), the values would be approximately 145% (nearly two-and-a-half times) larger.
d The PM2.5-related benefits (derived from benefit-per-ton values) presented in this table assume a 3% discount rate in the
valuation of premature mortality to account for a twenty-year segmented cessation lag. If a 7% discount rate had been used,
the values would be approximately 9% lower.
"Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected under
this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions, the value of these
reductions should not be interpreted as zero. Rather, the reductions in non-CO2GHGs will contribute to this rule's climate
benefits, as explained in Section III.F.2. The SCC TSD notes the difference between the social cost of non-CO2 emissions
and CO2 emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses.
'Section 7.5 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates range as
follows: for Average SCC at 5%: $5-$l6; for Average SCC at 3%: $21-$45; for Average SCC at 2.5%: $35-$65; and for
95th percentile SCC at 3%: $65-$136. Section 7.5 also presents these SCC estimates. Note that net present value of reduced
CO2 emissions is calculated differently than other benefits. The same discount rate used to discount the value of damages
from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer
to the SCC TSD for more detail.
8-27
-------
Regulatory Impact Analysis
Table 8-16. Quantified Net Benefits Associated with the Lifetimes of 2012-2016 Model Year Vehicles
(Millions of 2007 dollars; 3% Discount Rate)
Quantified Annual Costs
(excluding fuel savings)2
Quantified Annual Benefits at
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95thpercentileSCCat3%
2012MY
$4,900
2013MY
$8,000
2014MY
$10,300
each assumed SCC value"'0
$20,500 $30,100 $40,400
$21,800 $32,000 $42,800
$22,800 $33,500 $44,900
$25,200 $36,900 $49,300
Quantified Net Benefits at each assumed SCC value"'0
Avg SCC at 5% $15,600 $22,100
Avg SCC at 3% $16,900 $24,000
Avg SCC at 2.5% $17,900 $25,500
95thpercentileSCCat3% $20,300 $28,900
$30,100
$32,500
$34,600
$39,000
2015MY
$12,700
$57,400
$60,800
$63,600
$69,400
$44,700
$48,100
$50,900
$56,700
2016MY
$15,600
$78,700
$83,300
$87,100
$95,400
$63,100
$67,700
$71,500
$79,800
SUM
$51,500
$227,000
$240,200
$252,200
$276,200
$175,500
$188,700
$200,700
$224,700
a Quantified annual costs as shown here are the increased costs for new vehicles in each given model year. Since
those costs are assumed to occur in the given model year (i.e., not over a several year time span), the discount
rate does not affect the costs.
b Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected under
this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions, the value of these
reductions should not be interpreted as zero. Rather, the reductions in non-CO2GHGs will contribute to this rule's climate
benefits, as explained in Section III.F.2. The SCC TSD notes the difference between the social cost of non-CO2 emissions
and CO2 emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses.
c Section 7.5 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates range as
follows: for Average SCC at 5%: $5-$l6; for Average SCC at 3%: $21-$45; for Average SCC at 2.5%: $35-$65; and for
95th percentile SCC at 3%: $65-$136. Section 7.5 also presents these SCC estimates. Note that net present value of reduced
CO2 emissions is calculated differently than other benefits. The same discount rate used to discount the value of damages
from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer
to the SCC TSD for more detail.
8-28
-------
Other Economic and Social Impacts
Table 8-17 Quantified Net Benefits Associated with the Lifetimes of 2012-2016 Model Year Vehicles
(Millions of 2007 dollars; 7% Discount Rate)
Quantified Annual Costs
(excluding fuel savings)3
Quantified Annual Benefits at
Avg SCC at 5%
Avg SCC at 3%
Avg SCC at 2.5%
95th percentile SCC at 3%
2012MY
$4,900
2013MY
$8,000
each assumed SCC value"'0
$16,100 $23,800
$17,400 $25,700
$18,400 $27,200
$20,800 $30,600
Quantified Net Benefits at each assumed SCC valueb>c
Avg SCC at 5% $11,200 $15,800
Avg SCC at 3% $12,500 $17,700
Avg SCC at 2.5% $13,500 $19,200
95th percentile SCC at 3% $15,900 $22,600
2014MY
$10,300
$31,800
$34,200
$36,300
$40,700
$21,500
$23,900
$26,000
$30,400
2015MY
$12,700
$45,200
$48,600
$51,400
$57,200
$32,500
$35,900
$38,700
$44,500
2016MY
$15,600
$61,800
$66,400
$70,200
$78,500
$46,200
$50,800
$54,600
$62,900
SUM
$51,500
$178,500
$191,700
$203,700
$227,700
$127,000
$140,200
$152,200
$176,200
a Quantified annual costs as shown here are the increased costs for new vehicles in each given model year. Since
those costs are assumed to occur in the given model year (i.e., not over a several year time span), the discount
rate does not affect the costs.
b Monetized GHG benefits exclude the value of reductions in non-CO2 GHG emissions (HFC, CH4 and N2O) expected under
this final rule. Although EPA has not monetized the benefits of reductions in these non-CO2 emissions, the value of these
reductions should not be interpreted as zero. Rather, the reductions in non-CO2GHGs will contribute to this rule's climate
benefits, as explained in Section III.F.2. The SCC TSD notes the difference between the social cost of non-CO2 emissions
and CO2 emissions, and specifies a goal to develop methods to value non-CO2 emissions in future analyses.
c Section 7.5 notes that SCC increases over time. Corresponding to the years in this table, the SCC estimates range as
follows: for Average SCC at 5%: $5-$l6; for Average SCC at 3%: $21-$45; for Average SCC at 2.5%: $35-$65; and for
95th percentile SCC at 3%: $65-$136. Section 7.5 also presents these SCC estimates. Note that net present value of reduced
CO2 emissions is calculated differently than other benefits. The same discount rate used to discount the value of damages
from future emissions (SCC at 5, 3, 2.5 percent) is used to calculate net present value of SCC for internal consistency. Refer
to the SCC TSD for more detail.
8-29
-------
Regulatory Impact Analysis
References
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 (Docket EPA-HQ-
OAR-2009-0472-0007).
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
(Docket EPA-HQ-OAR-2009-0472-0008).
3 U.S. Department of Energy, 2008, "Average Price of a New Car, 1970-2006,"
http://wwwl.eere.energy.gov/vehiclesandfuels/facts/2008_fotw520.html, accessed April 23,
2009 (Docket EPA-HQ-OAR-2009-0472-0009).
4 Solheim, Mark, 2006"State Car Tax Rankings,"
http://www.kiplinger.com/features/archives/2006/04/cartax.html, accessed April 23, 2009
(Docket EPA-HQ-OAR-2009-0472-0010).
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 (Docket EPA-HQ-
OAR-2009-0472-0011).
6"Auto Loan Interest Rate", EPA-HQ-OAR-2009-0472-11575 http://auto-loan.interest.com,
1/22/10.
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 (Docket
EPA-HQ-OAR-2009-0472-0013).
8 Kleit A.N., 1990, "The Effect of Annual Changes in Automobile Fuel Economy Standards,"
Journal of Regulatory Economics 2: 151-172 (Docket EPA-HQ-OAR-2009-0472-0015);
McCarthy, Patrick S., 1996. "Market Price and Income Elasticities of New Vehicle
Demands." Review of Economics and Statistics 78: 543-547 (Docket EPA-HQ-OAR-2009-
0472-0016); Goldberg, Pinelopi K., 1998. "The Effects of the Corporate Average Fuel
Efficiency Standards in the US," Journal of Industrial Economics 46(1): 1-33 (Docket EPA-
HQ-OAR-2009-0472-0017); Greene, David L., "Feebates, Footprints and Highway Safety,"
Transportation Research Part D 14 (2009): 375-384 (Docket EPA-HQ-OAR-2009-0472-
0019).
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 (Docket EPA-HQ-OAR-2009-0472-0021); 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 (Docket EPA-HQ-OAR-2009-0472-0017).
8-30
-------
Other Economic and Social Impacts
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 (Docket EPA-HQ-OAR-2009-0472-0022); 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
(Docket EPA-HQ-OAR-2009-0472-0023).
11 McManus, Walter M., "Can Proactive Fuel Economy Strategies Help Automakers Mitigate
Fuel-Price Risks?" University of Michigan Transportation Research Institute, September 14,
2006 (Docket EPA-HQ-OAR-2009-0472-0024).
12 Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market
Equilibrium," Econometrica 63(4) (July 1995): 841-940 (Docket EPA-HQ-OAR-2009-0472-
0025); 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 (Docket EPA-HQ-OAR-2009-0472-0026).
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
(Docket EPA-HQ-OAR-2009-0472-0029).
14 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 (Docket EPA-HQ-OAR-2009-0472-0030).
15 Kleit, Andrew N., "Impacts of Long-Range Increases in the Fuel Economy (CAFE)
Standard," Economic Inquiry 42(2) (April 2004): 279-294 (Docket EPA-HQ-OAR-2009-
0472-0031).
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 (Docket EPA-HQ-OAR-2009-0472-0032).
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 (Docket EPA-HQ-OAR-2009-0472-0029); 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 (Docket EPA-HQ-OAR-
2009-0472-0030).
18 E.g., Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market
Equilibrium," Econometrica 63(4) (July 1995): 841-940 (Docket EPA-HQ-OAR-2009-0472-
0025).
8-31
-------
Regulatory Impact Analysis
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 (Docket EPA-HQ-OAR-2009-0472-0022).
20 NERA Economic Consulting, "Evaluation of NHTSA's Benefit-Cost Analysis of 2011-
2015 CAFE Standards," 2008, available at
http://www.heartland.org/policybot/results/23495/Evaluation_of_NHTSAs_BenefitCost_Anal
ysis_Of_20112015_CAFE_Standards.html (Docket EPA-HQ-OAR-2009-0472-0033).
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 (Docket EPA-HQ-OAR-2009-0472-0030).
22 E.g., Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market
Equilibrium," Econometrica 63(4) (July 1995): 841-940 (Docket EPA-HQ-OAR-2009-0472-
0025).
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 (Docket EPA-HQ-OAR-2009-0472-0029).
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
(Docket EPA-HQ-OAR-2009-0472-0029); 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 (Docket EPA-HQ-OAR-2009-0472-0034).
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 (Docket EPA-HQ-OAR-2009-0472-0035).
26 Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market
Equilibrium," Econometrica 63(4) (July 1995): 841-940 (Docket EPA-HQ-OAR-2009-0472-
0025).
27 Goldberg, Pinelopi Koujianou, "Product Differentiation and Oligopoly in International
Markets: The Case of the U.S. Automobile Industry," Econometrica 63(4) (July 1995): 891-
951 (Docket EPA-HQ-OAR-2009-0472-0021).
28 Brownstone, David, and Kenneth Train, "Forecasting New Product Penetration with
Flexible Substitution Patterns," Journal of Econometrics 89 (1999): 109-129 (Docket EPA-
HQ-OAR-2009-0472-0036); Brownstone, David, David S. Bunch, and Kenneth Train, "Joint
8-32
-------
Other Economic and Social Impacts
Mixed Logit Models of Stated and Revealed Preferences for Alternative-Fuel Vehicles,"
Transportation Research Part B 34 (2000): 315-338 (Docket EPA-HQ-OAR-2009-0472-
00357); 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 (Docket EPA-HQ-OAR-2009-0472-0038); 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 (Docket EPA-HQ-OAR-2009-0472-0039).
29 Petrin, Amil, "Quantifying the Benefits of New Products: The Case of the Minivan,"
Journal of Political Economy 110 (2002): 705-729 (Docket EPA-HQ-OAR-2009-0472-
0040); 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 (Docket EPA-HQ-OAR-2009-0472-0026).
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 (Docket EPA-HQ-OAR-2009-0472-0030).
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
(Docket EPA-HQ-OAR-2009-0472-0029); 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 (Docket EPA-HQ-OAR-2009-0472-0034).
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 (Docket EPA-HQ-OAR-2009-0472-0023); 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 (Docket EPA-HQ-OAR-2009-0472-0034); Greene, David L., "Feebates,
Footprints and Highway Safety," Transportation Research Part D (2009) (in press) (Docket
EPA-HQ-OAR-2009-0472-0019).
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 (Docket EPA-HQ-OAR-2009-0472-0032).
34 Turrentine, Thomas S., and Kenneth S. Kurani, "Car Buyers and Fuel Economy?" Energy
Policy 35 (2007): 1213-1223 (Docket EPA-HQ-OAR-2009-0472-0041).
35 E.g., Espey, Molly, and Santosh Nair, 2005. "Automobile Fuel Economy: What Is It
Worth?" Contemporary Economic Policy 23: 317-323 (Docket EPA-HQ-OAR-2009-0472-
0042).
8-33
-------
Regulatory Impact Analysis
36 Larrick, Richard P., and Jack B. Soil, 2008. "The MPG Illusion," Science 320(5883):
1593-1594 (Docket EPA-HQ-OAR-2009-0472-0043).
37 E.g., Goldberg, Pinelopi Koujianou, "Product Differentiation and Oligopoly in International
Markets: The Case of the U.S. Automobile Industry," Econometrica 63(4) (July 1995): 891-
951 (Docket EPA-HQ-OAR-2009-0472-0021).
38 E.g., Berry, Steven, James Levinsohn, and Ariel Pakes, "Automobile Prices in Market
Equilibrium," Econometrica 63(4) (July 1995): 841-940 (Docket EPA-HQ-OAR-2009-0472-
0025).
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) (Docket EPA-HQ-OAR-2009-0472-0044); Li, Shanjun, Christopher Timmins, and
Roger H. von Haefen, 2009. "How Do Gasoline Prices Affect Fleet Fuel Economy?"
American Economic Journal: Economic Policy 1(2): 113-137; Congressional Budget Office
(2008). Effects of Gasoline Prices on Driving Behavior and Vehicle Markets. The Congress
of the United States, Pub. No. 2883; West. Sarah (2007). "The Effect of Gasoline Prices on
the Demand for Sport Utility Vehicles." Working Paper. Macalester College,
http://www.macalester.edu/~wests/SarahWestMEA2007.pdf, accessed 1/25/10.
40 Greene, David L., and Jin-Tan Liu, 1988. "Automotive Fuel Economy Improvements and
Consumers' Surplus," Transportation Research Part A 22A (3): 203-218 (Docket EPA-HQ-
OAR-2009-0472-0045).
41 Greene, David L. (2010). "How Consumers Value Fuel Economy: A Literature Review."
EPA Report EPA-420-R-10-008. (Docket EPA-HQ-OAR-2009-0472-11575).
42 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 (Docket EPA-HQ-OAR-2009-0472-0046).
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 (Docket EPA-HQ-OAR-2009-0472-0046); 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 (Docket EPA-HQ-OAR-2009-0472-0054).
44 For instance, Kleit, Andrew N., "Impacts of Long-Range Increases in the Fuel Economy
(CAFE) Standard," Economic Inquiry 42(2) (April 2004): 279-294 (Docket EPA-HQ-OAR-
2009-0472-0031); 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 (Docket EPA-HQ-OAR-
2009-0472-0032); Klier, Thomas, and Joshua Linn, 2008. "New Vehicle Characteristics and
the Cost of the Corporate Average Fuel Economy Standard," Federal Reserve Bank of
8-34
-------
Other Economic and Social Impacts
Chicago WP 2008-13, at
http://www.chicagofed.org/publications/workingpapers7wp2008_13.pdf (accessed 5/6/09)
(Docket EPA-HQ-OAR-2009-0472-0047); 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) (Docket EPA-HQ-
OAR-2009-0472-0048).
45 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 (Docket EPA-HQ-OAR-2009-0472-0032).
46 IEA. 2007. "Mind the Gap: Quantifying Principal-Agent Problems in Energy Efficiency."
Paris, France: International Energy Agency (Docket EPA-HQ-OAR-2009-0472-0049); 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
(Docket EPA-HQ-OAR-2009-0472-0050); 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 (Docket EPA-
HQ-OAR-2009-0472-0051); Tietenberg, T. (2009). "Reflections - Energy Efficiency Policy:
Pipe Dream or Pipeline to the Future?" Review of Environmental Economics and Policy.
Vol.3, 2: 304-320 (Docket EPA-HQ-OAR-2009-0472-0052).
47 Helfand, Gloria, and Ann Wolverton (2009). "Evaluating the Consumer Response to Fuel
Economy: A Review of the Literature." U.S. Environmental Protection Agency, National
Center for Environmental Economics Working Paper 2009-4, available at
http://yosemite.epa.gov/EE/epa/eed.nsf/WPNumber/2009-047OpenDocument, accessed
1/25/10 (Docket EPA-HQ-OAR-2009-0472-11520).
48 Jaffe, A.B., and R.N Stavins (1994). "The Energy Paradox and the Diffusion of
Conservation Technology." Resource and Energy Economics 16(2): 91-122 (Docket EPA-
HQ-OAR-2009-0472-11415); Allcott, Hunt, and Nathan Wozny, "Gasoline Prices, Fuel
Economy, and the Energy Paradox," working paper, available at
http://web.mit.edu/allcott/www/Allcott%20and%20Wozny%202010%20-
%20Gasoline%20Prices,%20Fuel%20Economy,%20and%20the%20Energy%20Paradox.pdf
(Docket EPA-HQ-OAR-2009-0472-11554).
49 Barber, Brad, Terrance Odean, and Lu Zheng. "Out of Sight, Out of Mind: The Effects of
Expenses on Mutual Fund Flows." Journal of Business 78(6) (2005): 2095-2119 (Docket
EPA-HQ-OAR-2009-0472-11555)^
50 Welch, David. "CAFE: A Prize for Making Gas-Guzzlers?" Business Week April 16,
2009,
http://www.businessweek.com/magazine/content/09_17/b4128048030307.htm?chan=magazin
e+channel_what's+next, accessed 7/7/09 (Docket EPA-HQ-OAR-2009-0472-0053).
8-35
-------
Regulatory Impact Analysis
51 Berry, Steven, James Levinsohn, and Ariel Pakes (July 1995). "Automobile Prices in
Market Equilibrium," Econometrica 63(4): 841-940 (Docket EPA-HQ-OAR-2009-0472-
0025).
52 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 (Docket EPA-HQ-OAR-2009-0472-0023).
53 Rex, Emma, and Henrikke Baumann (2007). "Beyond ecolabels: what green marketing
can learn from conventional marketing." Journal of Cleaner Production 15: 567-576.
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 RIA in support of the "Corporate Average Fuel Economy Standards For
MY 2011 - 2015 Passenger Cars and Light Trucks", April 2008
57 See http://ostpxweb.dot.gov/policv/Data/VOT97guid.pdf and
http://ostpxweb.dot.gov/policv/Data/VOTrevisionl_2-1 l-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's Model 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/final/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).
8-36
-------
Other Economic and Social Impacts
Available at http://www.rff.org/rff/Documents/RFF-DP-02-12.pdf (last accessed July 27,
2009).
64 Interagency Working Group on Social Cost of Carbon, U.S. Government, with participation
by Council of Economic Advisers, Council on Environmental Quality, Department of
Agriculture, Department of Commerce, Department of Energy, Department of Transportation,
Environmental Protection Agency, National Economic Council, Office of Energy and Climate
Change, Office of Management and Budget, Office of Science and Technology Policy, and
Department of Treasury, "Social Cost of Carbon for Regulatory Impact Analysis Under
Executive Order 12866" February 2010, available in docket EPA-HQ-OAR-2009-0472.
8-37
-------
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. 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. This requirement does not apply if the agency
certifies that the rule will not have a significant economic impact on a substantial number
of small entities.
The following discussion 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 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 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 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 this assessment, EPA identified a total of about 47 vehicle entities, 33 of
which are vehicle manufacturers. Of a total of 33 manufacturers, two fit the SBA
9-1
-------
Regulatory Impact Analysis
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 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 exempting them
will have a negligible impact on the GHG emissions reductions from the standards.
EPA has not conducted a Regulatory Flexibility Analysis or a SBREFA SBAR
Panel for the rule because we are certifying that the rule would not have a significant
economic impact on a substantial number of small entities. EPA is exempting
manufacturers, domestic and foreign, meeting SBA's size definitions of small business as
described in 13 CFR 121.201. EPA will 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 exemption will
have a negligible impact on the GHG emissions reductions from the final standards.A
To ensure that EPA is aware of which companies would be exempt, EPA
proposed to require that such entities submit a declaration to EPA containing a detailed
written description of how that manufacturer qualifies as a small entity under the
provisions of 13 CFR 121.201. EPA has reconsidered the need for this additional
submission under the regulations and is deleting it as not necessary. We already have
A It should be noted that EPA is deferring CO2 standards for small volume manufacturers with annual sales
less than 5,000 vehicles. See preamble section III.B.6. This deferral is not dependent on whether the entity
in question meets the SBA definition of small entity.
9-2
-------
Small Business Flexibility Analysis
information on the limited number of small entities that we expect would receive the
benefits of the exemption, and do not need the proposed regulatory requirement to be
able to effectively implement this exemption for those parties who in fact meet its terms.
Small entities are currently covered by a number of EPA motor vehicle emission
regulations, and they routinely submit information and data on an annual basis as part of
their compliance responsibilities. The net effect is that these entities are not regulated by
the light duty vehicle greenhouse gas rule.
Responses to comments that the rule has an adverse impact on small entity
stationary sources which are not regulated by the rule can be found at Section 5.14 of the
Response to Comments.
9-3
------- |